xorg-x11-xauth-6.8.2-1.EL.13.16> A AlpA?C'/%xH tP3ุQb29l[u^,|a4ibjtb35e247d60442fdf1436b8f4b6baf0268361d761;Hyv_GT?C'/%xOl!NAʓ!AﰏJr ԋ ^`,[>>?d " ; $4 < D T  8Hh  B X\d(p8x9:BϺFGHIXY$Z8[@\D]T^|bСdefkltuvwxyCxorg-x11-xauth6.8.21.EL.13.16X authority file utilityThe xauth program is used to edit and display the authorization information used in connecting to the X server.C'#yort.fnal.govRScientific LinuxMIT/X11, and othersUser Interface/Xhttp://xorg.freedesktop.orglinuxi386*k|큤$C'#C'#C'#C'#b95a54b532c807be09c06587d06864f3e3e9fd99e32e9d65c8120475056db48dca36c367e35d45ed052e22967b4f23e3048e13911f1580f70f1532c29a0229dcrootrootrootrootrootrootrootrootxorg-x11-6.8.2-1.EL.13.16.src.rpmmkxauthxauthxorg-x11-xauth@@@@@@@@JJ/bin/shlibX11.so.6libXext.so.6libXmuu.so.1libc.so.6libc.so.6(GLIBC_2.0)libc.so.6(GLIBC_2.1)libc.so.6(GLIBC_2.3)rpmlib(CompressedFileNames)rpmlib(PayloadFilesHavePrefix)3.0.4-14.0-1XFree864.2.0-50.114.3.3CB?B?BvBvBfBdJBIBBB8ȀB7wB6%B21B21B0߀B0߀B.?>?l???T??w?>+>6>6>M>M>Ҫ>ζ>>L>L>>>>>>z>z>z>׀>4>>>>>>>rm>j>j>i2>b>^>^>]U>\>Z>Z>Z>Y`>V>V>V>V>Ul>Ul>R>R>Qw>Qw>M>M>L1>L1>L1>L1>I>I>H=>H=>B>A>A>@T>@T>?><_><_>9>9>9>8k>1Ӏ>1Ӏ>1Ӏ>/0>->,>+<>)>)>%>%>%>$> >j>j>>$>====t==.=.==:=====E===\= = =⹀=======s=s=Պ=і=і=і=D====P==[= =g=g=g==8=8======O====Z=====q= ===̀===~={={=z=z=z=w=uM=r=n=n=n=j=io=e{=e{=a=\@=Z=Y=XL=S=Q=Pc=O=O=Ln=Ln=K=Hz=G(=A=A=A=??===<=<=9=7V=*'==a=a==l=ɀ=x=x= ====<<<98J9+9)9I999`99888Ӏ88捀88$8́Soren Sandmann 6.8.2-1.EL.13.16Mike A. Harris 6.8.2-1.EL.13.6Mike A. Harris 6.8.2-1.EL.13.5Mike A. Harris 6.8.2-1.EL.13.4Kristian Høgsberg 6.8.2-1.EL.13.3Kristian Høgsberg 6.8.2-1.EL.13.2Kristian Høgsberg 6.8.2-1.EL.13.1Kristian Høgsberg 6.8.2-1.EL.13Kristian Høgsberg 6.8.2-13Soeren Sandmann 6.8.2-12Mike A. Harris 6.8.2-11Mike A. Harris Mike A. Harris 6.8.2-1.FC3.10testMike A. Harris 6.8.2-10Mike A. Harris 6.8.2-9Mike A. Harris 6.8.2-1.FC3.8testSoeren Sandmann 6.8.2-8Soeren Sandmann 6.8.2-7Mike A. Harris 6.8.2-1.FC3.6testMike A. Harris 6.8.2-6Mike A. Harris 6.8.2-1.EL.5Mike A. Harris 6.8.2-5Soeren Sandmann 6.8.2-4Soeren Sandmann 6.8.2-3Mike A. Harris 6.8.2-2Mike A. Harris 6.8.2-1Mike A. Harris 6.8.1.904-2Mike A. Harris 6.8.1.904-1Mike A. Harris 6.8.1.903-4Soeren Sandmann 6.8.1.903-3Mike A. Harris 6.8.1.903-2Mike A. Harris 6.8.1.903-1Mike A. Harris 6.8.1.902-7Mike A. Harris 6.8.1.902-6Mike A. Harris 6.8.1.902-5Mike A. Harris 6.8.1.902-4Mike A. Harris 6.8.1.902-3Mike A. Harris 6.8.1.902-2Mike A. Harris 6.8.1.902-1Mike A. Harris 6.8.1.901-1Mike A. Harris 6.8.1-25Kristian Høgsberg Kristian Høgsberg 6.8.1-24Kristian Høgsberg Soeren Sandmann 6.8.1-22.ELSoeren Sandmann Mike A. Harris Kristian Høgsberg Kristian Høgsberg 6.8.1-22Mike A. Harris 6.8.1-21.ELMike A. Harris 6.8.1-21Mike A. Harris 6.8.1-20Kristian Høgsberg Mike A. Harris 6.8.1-19.with_fonts.0Mike A. Harris 6.8.1-19Mike A. Harris 6.8.1-18Kristian Høgsberg 6.8.1-17Kristian Høgsberg 6.8.1-16Kristian Høgsberg 6.8.1-15Mike A. Harris 6.8.1-14Kristian Høgsberg Kristian Høgsberg 6.8.1-13Kristian Høgsberg Kristian Høgsberg Kristian Høgsberg Kristian Høgsberg 6.8.1-12Mike A. Harris 6.8.1-11Mike A. Harris 6.8.1-10 6.8.1-9Mike A. Harris Mike A. Harris 6.8.1-8Soren Sandmann 6.8.1-7 6.8.1-6Mike A. Harris 6.8.1-5Mike A. Harris Mike A. Harris 6.8.1-4Mike A. Harris 6.8.1-3Mike A. Harris 6.8.1-2Kristian Høgsberg 6.8.1-1Mike A. Harris 6.8.0-5Bill Nottingham 6.8.0-4Kristian Høgsberg 6.8.0-3Kristian Høgsberg 6.8.0-2Mike A. Harris 6.8.0-1Mike A. Harris Mike A. Harris 6.7.99.903-6Bill Nottingham 6.7.99.903-5Mike A. Harris 6.7.99.903-4Mike A. Harris 6.7.99.903-3Mike A. Harris 6.7.99.903-2Mike A. Harris 6.7.99.903-1Mike A. Harris 6.7.99.902-8Mike A. Harris 6.7.99.902-7Mike A. Harris 6.7.99.902-6Mike A. Harris 6.7.99.902-5Mike A. Harris 6.7.99.902-4Mike A. Harris 6.7.99.902-3Mike A. Harris 6.7.99.902-2Mike A. Harris 6.7.99.902-1Mike A. Harris 6.7.99.901-2Mike A. Harris 6.7.99.901-1Mike A. Harris 6.7.99.2-5Jeremy Katz - 6.7.99.2-4Mike A. Harris 6.7.99.2-3Mike A. Harris 6.7.99.2-2Mike A. Harris 6.7.99.2-1Mike A. Harris 6.7.99.1-1Mike A. Harris 6.7.99.0-0.2004_08_05.1Mike A. Harris 6.7.99.0-0.2004_08_05.0Mike A. Harris 6.7.0-7.2Mike A. Harris 6.7.0-7.1Mike A. Harris 6.7.0-7Mike A. Harris 6.7.0-6.2Mike A. Harris 6.7.0-6.1Mike A. Harris Mike A. Harris 6.7.0-6Mike A. Harris 6.7.0-5.1Mike A. Harris 6.7.0-5Elliot Lee 6.7.0-4Mike A. Harris 6.7.0-3Mike A. Harris 6.7.0-2Mike A. Harris 6.7.0-1John Dennis 6.7.0-0.5Mike A. Harris 6.7.0-0.4Mike A. Harris 6.7.0-0.3Mike A. Harris 6.7.0-0.2Mike A. Harris 6.7.0-0.1Mike A. Harris Mike A. Harris 0.6.6-0.2004_03_30.5Mike A. Harris 0.6.6-0.2004_03_30.4Mike A. Harris Mike A. Harris 0.6.6-0.2004_03_30.3Mike A. Harris 0.6.6-0.2004_03_30.2Mike A. Harris 0.6.6-0.2004_03_30.1Mike A. Harris 0.6.6-0.2004_03_30.0Mike A. Harris 0.0.6.6-0.0.2004_03_11.11John Dennis 0.0.6.6-0.0.2004_03_11.10Mike A. Harris Mike A. Harris 0.0.6.6-0.0.2004_03_11.9Mike A. Harris 0.0.6.6-0.0.2004_03_11.8Mike A. Harris 0.0.6.6-0.0.2004_03_11.7Mike A. Harris 0.0.6.6-0.0.2004_03_11.6Mike A. Harris 0.0.6.6-0.0.2004_03_11.5Mike A. Harris 0.0.6.6-0.0.2004_03_11.4Mike A. Harris 0.0.6.6-0.0.2004_03_11.3Mike A. Harris 0.0.6.6-0.0.2004_03_11.2Mike A. Harris 0.0.6.6-0.0.2004_03_11.1Mike A. Harris 0.0.6.6-0.0.2004_03_11.0Mike A. Harris Mike A. Harris 0.0.6.6-0.0.2004_03_09.1Mike A. Harris 0.0.6.6-0.0.2004_03_09.0Mike A. Harris Mike A. Harris Mike A. Harris 4.3.0-62.ELMike A. Harris 4.3.0-62Mike A. Harris 4.3.0-61Mike A. Harris 4.3.0-60Mike A. Harris 4.3.0-59Mike A. Harris 4.3.0-58Mike A. Harris 4.3.0-57Mike A. Harris 4.3.0-56Mike A. Harris 4.3.0-55Mike A. Harris 4.3.0-54Mike A. Harris 4.3.0-53Mike A. Harris 4.3.0-52Mike A. Harris 4.3.0-51Mike A. Harris 4.3.0-50Mike A. Harris 4.3.0-2.90.49Mike A. Harris 4.3.0-49.ELMike A. Harris 4.3.0-49Mike A. Harris 4.3.0-45.0.2.EL.testMike A. Harris 4.3.0-45.0.2Mike A. Harris 4.3.0-45.0.1.EL.testMike A. Harris 4.3.0-45.0.1Mike A. Harris 4.3.0-45Mike A. Harris 4.3.0-44.ELMike A. Harris 4.3.0-44Mike A. Harris 4.3.0-43.1Mike A. Harris 4.3.0-2.90.43Mike A. Harris 4.3.0-43Mike A. Harris 4.3.0-42.2Mike A. Harris 4.3.0-42Mike A. Harris 4.3.0-41Mike A. Harris 4.3.0-40Mike A. Harris 4.3.0-39Mike A. Harris 4.3.0-38.1Mike A. Harris 4.3.0-38Mike A. Harris 4.3.0-37Mike A. Harris 4.3.0-36.2Mike A. Harris 4.3.0-36Mike A. Harris 4.3.0-35.ELMike A. Harris 4.3.0-35Mike A. Harris 4.3.0-34.2Mike A. Harris 4.3.0-34.1Mike A. Harris 4.3.0-34Mike A. Harris 4.3.0-33.1Mike A. Harris 4.3.0-33Mike A. Harris 4.3.0-32.ELMike A. Harris 4.3.0-32Mike A. Harris 4.3.0-31Mike A. Harris 4.3.0-30.ELMike A. Harris 4.3.0-30Mike A. Harris 4.3.0-29.1Mike A. Harris 4.3.0-29Karsten Hopp 4.3.0-28Mike A. Harris 4.3.0-27Mike A. Harris 4.3.0-26Mike A. Harris 4.3.0-25Mike A. Harris 4.3.0-24Mike A. Harris 4.3.0-23.dbgMike A. Harris 4.3.0-22.4Mike A. Harris 4.3.0-22.3Mike A. Harris 4.3.0-22.2Mike A. Harris 4.3.0-22.1Mike A. Harris 4.3.0-22Mike A. Harris 4.3.0-21Mike A. Harris 4.3.0-20Mike A. Harris 4.3.0-19Mike A. Harris 4.3.0-18Mike A. Harris 4.3.0-17.1Mike A. Harris 4.3.0-17Mike A. Harris 4.3.0-16.ELMike A. Harris 4.3.0-16Mike A. Harris 4.3.0-15.1Mike A. Harris 4.3.0-15Mike A. Harris 4.3.0-14.ELMike A. Harris 4.3.0-14Mike A. Harris 4.3.0-13.1Mike A. Harris 4.3.0-13Mike A. Harris 4.3.0-12Mike A. Harris 4.3.0-11.2Mike A. Harris 4.3.0-11Mike A. Harris 4.3.0-10.1.ELBill Nottingham 4.3.0-10Mike A. Harris 4.3.0-9Mike A. Harris 4.3.0-8.3Mike A. Harris 4.3.0-8.RHEL.1Mike A. Harris 4.3.0-8Mike A. Harris 4.3.0-7.el.1Mike A. Harris 4.3.0-7.el.0Mike A. Harris 4.3.0-6Mike A. Harris 4.3.0-5Mike A. Harris 4.3.0-4Mike A. Harris 4.3.0-3.9Mike A. Harris 4.3.0-3.8Mike A. Harris 4.3.0-3.7Mike A. Harris 4.3.0-3.6Mike A. Harris 4.3.0-3.5Mike A. Harris 4.3.0-3.4Mike A. Harris 4.3.0-3.3Mike A. Harris 4.3.0-3.2Mike A. Harris 4.3.0-3.1Mike A. Harris 4.3.0-3Mike A. Harris 4.3.0-2.2Mike A. Harris 4.3.0-2.1Mike A. Harris 4.3.0-2Mike A. Harris 4.3.0-1Mike A. Harris 4.2.99.902-20030226.0Mike A. Harris 4.2.99.902-20030225.0Mike A. Harris 4.2.99.902-20030224.2Mike A. Harris 4.2.99.902-20030224.1Mike A. Harris 4.2.99.902-20030224.0Mike A. Harris 4.2.99.902-20030223.0Mike A. Harris 4.2.99.902-20030220.3Mike A. Harris 4.2.99.902-20030220.2Mike A. Harris 4.2.99.902-20030220.1Mike A. Harris 4.2.99.902-20030220.0Mike A. Harris 4.2.99.902-20030218.3Mike A. Harris 4.2.99.902-20030218.2Mike A. Harris 4.2.99.902-20030218.1Mike A. Harris 4.2.99.902-20030218.0Mike A. Harris 4.2.99.902-20030217.1Mike A. Harris 4.2.99.902-20030217.0Mike A. Harris 4.2.99.901-20030213.1Mike A. Harris 4.2.99.901-20030213.0Mike A. Harris 4.2.99.901-20030212.0Mike A. Harris 4.2.99.901-20030211.3Mike A. Harris 4.2.99.901-20030211.2Mike A. Harris 4.2.99.901-20030211.1Mike A. Harris 4.2.99.901-20030211.0Mike A. Harris 4.2.99.901-20030209.2Mike A. Harris 4.2.99.901-20030209.1Mike A. Harris 4.2.99.901-20030209.0Mike A. Harris 4.2.99.901-20030206.0Mike A. Harris 4.2.99.901-20030205.0Mike A. Harris 4.2.99.901-20030203.2Mike A. Harris 4.2.99.901-20030203.1Mike A. Harris 4.2.99.901-20030203.0Mike A. Harris 4.2.99.4-20030129.5Mike A. Harris 4.2.99.4-20030129.4Mike A. Harris 4.2.99.4-20030129.3Mike A. Harris 4.2.99.4-20030129.2Mike A. Harris 4.2.99.4-20030129.1Mike A. Harris 4.2.99.4-20030129.0Mike A. Harris 4.2.99.4-20030121.6Mike A. Harris 4.2.99.4-20030121.4Mike A. Harris 4.2.99.4-20030121.3Mike A. Harris 4.2.99.4-20030121.2Mike A. Harris 4.2.99.4-20030121.1Mike A. Harris 4.2.99.4-20030121.0Mike A. Harris 4.2.99.3-20030118.3Mike A. Harris 4.2.99.3-20030118.2Mike A. Harris 4.2.99.3-20030118.1Mike A. Harris 4.2.99.3-20030118.0Mike A. Harris 4.2.99.3-20030115.1Mike A. Harris 4.2.99.3-20030115.0Mike A. Harris 4.2.99.3-20021230.7Mike A. Harris 4.2.99.3-20021230.6Mike A. Harris 4.2.99.3-20021230.5Mike A. Harris 4.2.99.3-20021230.4Mike A. Harris 4.2.99.3-20021230.3Mike A. Harris 4.2.99.3-20021230.2Mike A. Harris 4.2.99.3-20021230.1Mike A. Harris 4.2.99.3-20021230.0Mike A. Harris 4.2.99.2-0.20021217.1Mike A. Harris 4.2.99.2-0.20021217.0Mike A. Harris 4.2.99.2-0.20021215.0Mike A. Harris 4.2.99.2-0.20021210.1Mike A. Harris 4.2.99.2-0.20021210.0Mike A. Harris 4.2.99.2-0.20021209.0Mike A. Harris 4.2.99.2-0.20021208.1Mike A. Harris 4.2.99.2-0.20021208.0Mike A. Harris 4.2.99.2-0.20021126.10Mike A. Harris 4.2.99.2-0.20021126.9Mike A. Harris 4.2.99.2-0.20021126.8Mike A. Harris 4.2.99.2-0.20021126.7Mike A. Harris 4.2.99.2-0.20021126.6Mike A. Harris 4.2.99.2-0.20021126.5Mike A. Harris 4.2.99.2-0.20021126.3Mike A. Harris 4.2.99.2-0.20021126.2Mike A. Harris 4.2.99.2-0.20021126.1axpMike A. Harris 4.2.99.2-0.20021126.1Mike A. Harris 4.2.99.2-0.20021126.0Mike A. Harris 4.2.99.2-0.20021122.4Mike A. Harris 4.2.99.2-0.20021122.3Mike A. Harris 4.2.99.2-0.20021122.2Mike A. Harris 4.2.99.2-0.20021122.1Mike A. Harris 4.2.99.2-0.20021122.0Mike A. Harris 4.2.99.2-0.20021110.11Mike A. Harris 4.2.99.2-0.20021110.10Mike A. Harris 4.2.99.2-0.20021110.9Mike A. Harris 4.2.99.2-0.20021110.8Mike A. Harris 4.2.99.2-0.20021110.7Mike A. Harris 4.2.99.2-0.20021110.6Mike A. Harris 4.2.99.2-0.20021110.5Mike A. Harris 4.2.99.2-0.20021110.4Mike A. Harris 4.2.99.2-0.20021110.3Mike A. Harris 4.2.99.2-0.20021110.2Mike A. Harris 4.2.99.2-0.20021110.1Mike A. Harris 4.2.99.2-0.20021110.0Mike A. Harris 4.2.99.2-0.20021105.2Mike A. Harris 4.2.99.2-0.20021105.1Mike A. Harris 4.2.99.2-0.20021105.0Mike A. Harris 4.2.99.2-0.20021101.6Mike A. Harris 4.2.99.2-0.20021101.5Mike A. Harris 4.2.99.2-0.20021101.4Mike A. Harris 4.2.99.2-0.20021101.2Mike A. Harris 4.2.99.2-0.20021101.1Mike A. Harris 4.2.99.2-0.20021023.2Mike A. Harris 4.2.99.2-0.20021023.0Mike A. Harris 4.2.99.1-0.20021018.6Mike A. Harris 4.2.99.1-0.20021018.5Mike A. Harris 4.2.99.1-0.20021018.4Mike A. Harris 4.2.99.1-0.20021018.1Mike A. Harris 4.2.99.1-0.20021018.0Mike A. Harris 4.2.99.1-0.20021014.4Mike A. Harris 4.2.99.1-0.20021014.3Mike A. Harris 4.2.99.1-0.20021014.2Mike A. Harris 4.2.99.1-0.20021014.1Mike A. Harris 4.2.99.1-0.20021014.0Mike A. Harris 4.2.99.1-0.20021012.0Mike A. Harris 4.2.99.1-0.20021008.8Mike A. Harris 4.2.99.1-0.20021008.5Mike A. Harris 4.2.99.1-0.20021008.0Mike A. Harris 4.2.1-8Mike A. Harris 4.2.1-7Mike A. Harris 4.2.1-6.2Mike A. Harris 4.2.1-6.1Mike A. Harris 4.2.1-6Mike A. Harris 4.2.1-5.1Mike A. Harris 4.2.1-5Mike A. Harris 4.2.1-4Mike A. Harris 4.2.1-3.3Mike A. Harris 4.2.1-3.2Mike A. Harris 4.2.1-3Mike A. Harris 4.2.1-2Mike A. Harris 4.2.1-1Owen Taylor Mike A. Harris 4.2.0-71Mike A. Harris 4.2.0-70Mike A. Harris 4.2.0-69.1Mike A. Harris 4.2.0-69Mike A. Harris 4.2.0-68Mike A. Harris 4.2.0-67.6Mike A. Harris 4.2.0-67.5Mike A. Harris 4.2.0-67.1Mike A. Harris 4.2.0-67Mike A. Harris 4.2.0-66Mike A. Harris 4.2.0-65Mike A. Harris 4.2.0-64Mike A. Harris 4.2.0-63.1Mike A. Harris 4.2.0-63Mike A. Harris 4.2.0-60.4Mike A. Harris 4.2.0-60.3Mike A. Harris 4.2.0-60.2Mike A. Harris 4.2.0-60.1Mike A. Harris 4.2.0-60Mike A. Harris 4.2.0-59.2Mike A. Harris 4.2.0-59.1Mike A. Harris 4.2.0-59Mike A. Harris 4.2.0-58.1Mike A. Harris 4.2.0-58Mike A. Harris 4.2.0-57.1Mike A. Harris 4.2.0-57Mike A. Harris 4.2.0-56Mike A. Harris 4.2.0-55.2Mike A. Harris 4.2.0-55.1Mike A. Harris 4.2.0-55Tim Powers 4.2.0-54Mike A. Harris 4.2.0-53.1Mike A. Harris 4.2.0-53Elliot Lee 4.2.0-52.01Mike A. Harris 4.2.0-52Mike A. Harris 4.2.0-51Mike A. Harris 4.2.0-50.24Mike A. Harris 4.2.0-50.23Mike A. Harris 4.2.0-50.22Mike A. Harris 4.2.0-50.21Mike A. Harris 4.2.0-50.20Mike A. Harris 4.2.0-50.19Mike A. Harris 4.2.0-50.18Mike A. Harris 4.2.0-50.17Mike A. Harris 4.2.0-50.16Mike A. Harris 4.2.0-50.15Mike A. Harris 4.2.0-50.14Mike A. Harris 4.2.0-50.13Mike A. Harris 4.2.0-50.12Mike A. Harris 4.2.0-50.11Mike A. Harris 4.2.0-50.10Mike A. Harris 4.2.0-50.9Mike A. Harris 4.2.0-50.8Mike A. Harris 4.2.0-50.6Mike A. Harris 4.2.0-50.1Mike A. Harris 4.2.0-50Mike A. Harris 4.2.0-9.13Mike A. Harris 4.2.0-9.12Mike A. Harris 4.2.0-9.11Mike A. Harris 4.2.0-9.10Mike A. Harris 4.2.0-9.8Mike A. Harris 4.2.0-9.0Mike A. Harris 4.2.0-8.1.betaMike A. Harris 4.2.0-8Mike A. Harris 4.2.0-7Mike A. Harris 4.2.0-6.666Mike A. Harris 4.2.0-6.62Mike A. Harris 4.2.0-6.61Mike A. Harris Mike A. Harris 4.2.0-6.60Mike A. Harris Mike A. Harris 4.2.0-6.52Mike A. Harris 4.2.0-6.51Mike A. Harris 4.2.0-6.50Mike A. Harris 4.2.0-6.48Mike A. Harris 4.2.0-6.47Mike A. Harris 4.2.0-6.46Mike A. Harris 4.2.0-6.45Mike A. Harris 4.2.0-6.44Mike A. Harris 4.2.0-6.42Mike A. Harris 4.2.0-6.41Mike A. Harris 4.2.0-6.40Mike A. Harris 4.2.0-6.39Mike A. Harris 4.2.0-6.38Mike A. Harris 4.2.0-6.37Mike A. Harris 4.2.0-6.36Mike A. Harris 4.2.0-6.35Mike A. Harris 4.2.0-6.34Mike A. Harris 4.2.0-6.33Mike A. Harris 4.2.0-6.32Mike A. Harris 4.2.0-6.31Mike A. Harris 4.2.0-6.30Mike A. Harris 4.2.0-6.24Mike A. Harris 4.2.0-6.23Mike A. Harris 4.2.0-6.21Mike A. Harris 4.2.0-6.20Mike A. Harris 4.2.0-6.15Mike A. Harris 4.2.0-6.14Mike A. Harris 4.2.0-6Mike A. Harris 4.2.0-5.2Mike A. Harris 4.2.0-5.1Mike A. Harris 4.2.0-5Mike A. Harris 4.2.0-4.1Mike A. Harris 4.2.0-4.0Mike A. Harris 4.2.0-3.6Mike A. Harris 4.2.0-3.4Mike A. Harris 4.2.0-3.2Mike A. Harris 4.2.0-3.1Mike A. Harris 4.2.0-3.0Mike A. Harris 4.2.0-2.5Mike A. Harris 4.2.0-2.4Mike A. Harris 4.2.0-2.2Bernhard Rosenkraenzer 4.2.0-2.1Mike A. Harris 4.2.0-2.0Mike A. Harris 4.2.0-0.3Mike A. Harris 4.2.0-0.2Mike A. Harris 4.2.0-0.1Mike A. Harris 4.1.99.7-0.1Mike A. Harris 4.1.99.4-0.1Mike A. Harris 4.1.99.3-0.1Mike A. Harris 4.1.99.2-0.2.4Mike A. Harris 4.1.99.2-0.2.3Mike A. Harris 4.1.99.2-0.1.7Mike A. Harris 4.1.99.2-0.1.5Mike A. Harris 4.1.99.2-0.1.3Mike A. Harris 4.1.99.2-0.1.2Mike A. Harris 4.1.99.2-0.1.1Mike A. Harris 4.1.99.2-0.0.4Mike A. Harris 4.1.99.2-0.0.2Than Ngo Mike A. Harris 4.1.0-5Mike A. Harris 4.1.0-4Mike A. Harris Mike A. Harris 4.1.0-3.9Mike A. Harris 4.1.0-3.8Mike A. Harris 4.1.0-3.5Mike A. Harris Mike A. Harris 4.1.0-3 Mike A. Harris Mike A. Harris 4.1.0-2Mike A. Harris 4.1.0-1.1Mike A. Harris 4.1.0-1Mike A. Harris 4.1.0-0.99.2Mike A. Harris 4.1.0-0.99.1Mike A. Harris 4.1.0-0.20.5Mike A. Harris Mike A. Harris 4.1.0-0.20.1Mike A. Harris Mike A. Harris 4.1.0-0.20.0Mike A. Harris 4.1.0-0.9.13Mike A. Harris 4.1.0-0.9.12Mike A. Harris Bill Nottingham 4.1.0-0.9.11Mike A. Harris 4.1.0-0.9.10Mike A. Harris 4.1.0-0.9.9Mike A. Harris Mike A. Harris 4.1.0-0.9.8Mike A. Harris 4.1.0-0.9.7Mike A. Harris 4.1.0-0.9.4Mike A. Harris 4.1.0-0.9.3Mike A. Harris 4.1.0-0.9.2Mike A. Harris Mike A. Harris 4.1.0-0.9.1Mike A. Harris Mike A. Harris 4.1.0-0.9.0Mike A. Harris Mike A. Harris Tim Powers 4.1.0-0.8.6Mike A. Harris 4.1.0-0.8.5Mike A. Harris 4.1.0-0.8.1Mike A. Harris 4.1.0-0.7.51Mike A. Harris 4.1.0-0.7.50Mike A. Harris 4.1.0-0.7.12Mike A. Harris 4.1.0-0.7.8Mike A. Harris 4.1.0-0.7.6Mike A. Harris 4.1.0-0.7.5Mike A. Harris Mike A. Harris 4.1.0-0.6.1Mike A. Harris 4.1.0-0.6.0Mike A. Harris 4.1.0-0.5.6Mike A. Harris 4.1.0-0.5.6Mike A. Harris Mike A. Harris 4.1.0-0.1.5Mike A. Harris 4.1.0-0.0.2Mike A. Harris 4.1.0-0.0.1Mike A. Harris 4.0.99.901-0.0.6Mike A. Harris 4.0.99.901-0.0.4Mike A. Harris 4.0.99.900-0.2.5Mike A. Harris 4.0.99.900-0.2.2Bernhard Rosenkraenzer 4.0.99.900-0.2.1Mike A. Harris Mike A. Harris Bernhard Rosenkraenzer Mike A. Harris 4.0.99.3-0Mike A. Harris 4.0.3-14Mike A. Harris 4.0.3-13Mike A. Harris 4.0.3-12Mike A. Harris 4.0.3-11Mike A. Harris Mike A. Harris Mike A. Harris 4.0.3-10Mike A. Harris 4.0.3-9Mike A. Harris Mike A. Harris Mike A. Harris 4.0.3-8Mike A. Harris 4.0.3-7Mike A. Harris 4.0.3-6Mike A. Harris Mike A. Harris 4.0.3-5Mike A. Harris 4.0.3-4Mike A. Harris Mike A. Harris Mike A. Harris 4.0.3-3Mike A. Harris Mike A. Harris Mike A. Harris 4.0.3-2Mike A. Harris Mike A. Harris 4.0.3-1Mike A. Harris Mike A. Harris 4.0.2a-1Mike A. Harris 4.0.2a-1Mike A. Harris Mike A. Harris Bill Nottingham 4.0.2-12.1Mike A. Harris 4.0.2-12Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris 4.0.2-11Mike A. Harris 4.0.2-10Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Bill Nottingham Matt Wilson Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Mike A. Harris Bernhard Rosenkraenzer Bill Nottingham Bill Nottingham Bill Nottingham Bernhard Rosenkraenzer Mike A. Harris Bill Nottingham Bill Nottingham Mike A. Harris Bill Nottingham Mike A. Harris Mike A. Harris Mike A. Harris Bernhard Rosenkraenzer Bernhard Rosenkraenzer Bill Nottingham Bill Nottingham Bill Nottingham Bill Nottingham Bill Nottingham Bill Nottingham Bill Nottingham Bill Nottingham Bill Nottingham Bill Nottingham Bill Nottingham Matt Wilson Elliot Lee Bill Nottingham Bill Nottingham Bill Nottingham Bill Nottingham Elliot Lee Preston Brown Elliot Lee Bill Nottingham Bill Nottingham Bill Nottingham Bill Nottingham Bill Nottingham Bill Nottingham Bill Nottingham Preston Brown Bill Nottingham Jakub Jelinek Bill Nottingham Bernhard Rosenkraenzer Bill Nottingham Matt Wilson Bernhard Rosenkraenzer Bernhard Rosenkraenzer Bill Nottingham Nalin Dahyabhai Jakub Jelinek Jakub Jelinek Bill Nottingham Bill Nottingham Bill Nottingham Bill Nottingham Bill Nottingham Bernhard Rosenkraenzer Bernhard Rosenkraenzer Bernhard Rosenkraenzer Bernhard Rosenkraenzer Jakub Jelinek Bernhard Rosenkraenzer Bernhard Rosenkraenzer Bernhard Rosenkraenzer Bernhard Rosenkraenzer - Add security-CAN-2005-2495.patch to fix integer overflow in pixmap creation code.- Temporarily disable xorg-x11-6.8.2-use-linux-native-pciscan-by-default.patch which was added for a test build in 6.8.2-1.EL.13.4, as it is intended for RHEL4_U2, not RHEL4_U1. We will re-enable this for U2 once U1 is released.- Added xorg-x11-6.8.2-ati-ragexl-ia64-avoidcpiofix.patch to workaround issue on ia64 with CPIO disabled in ati Mach64 driver (#155609,155610). For future reference, this is also included in FC4 build 6.8.2-31.- Added xorg-x11-6.8.2-use-linux-native-pciscan-by-default.patch to fix PCI config space contention issue by changing the X server to default to using Linux native PCI interfaces instead of directly banging on PCI space itself (#146386,152608). Patch is from build 6.8.2-16 in FC4.- Dont install Xprint man page- Remove CVS directories from upstream tarball.- Add xorg-x11-6.8.1-ati-radeon-dynamic-clocks-fix-2.patch to revert radeon dynamic clock setup to what we had in 6.8.1. The 6.8.2 code still causes lockups on some systems (#152648).- Rebuild 6.8.2-13 as 6.8.2-1.EL.13 for RHEL 4 security errata.- Add XFree86-4.1.0-xpm-security-fix-CAN-2005-0605.patch (#150040).- Add xorg-x11-6.8.2-cursor-flicker.patch (#144022) - Fix xorg-x11-6.8.2-XScreenSaverQueryInfo-crash-fix.patch so it will apply- Added xorg-x11-6.8.2-config-StaticNeedsPicForShared.patch for (#108026) - Removed StaticNeedsPicForShared from host.def section - Added xorg-x11-6.8.2-XScreenSaverQueryInfo-crash-fix.patch to fix (#147890)- Search and destroy any CVS directories found in RPM_BUILD_DIR during rpm spec file build section prior to building. This fixes a problem where the 6.8.2 release was tarballed officially from a CVS checkout rather than a CVS export, causing the CVS metadata to be present, and some of it ends up getting installed in the final OS.- Rebuild 6.8.2-10 as 6.8.2-1.FC3.10test for Fedora Core 3 testing release- Re-enable xorg-x11-6.8.2-xnest-shape-fix.patch which has been updated now for 6.8.2 by Mark McLoughlin (#148763,138482)- Disable xorg-x11-6.8.2-xnest-shape-fix.patch added in 6.8.2-8, as it does not compile.- Rebuild 6.8.2-8 as 6.8.2-1.FC3.8test for Fedora Core 3 testing release- Add Patch9329: xorg-x11-6.8.2-xnest-shape-fix.patch (Patch by Mark McLoughlin). (#148763)- Added xorg-x11-6.8.2-fix-font-crash.patch to fix crash on fonts with NULL bits (#145546)- Rebuild 6.8.2-6 as 6.8.2-1.FC3.6test for Fedora Core 3 testing release- Added xorg-x11-6.8.2-ati-radeon-gcc4-fix.patch to fix bug in radeon driver when building with gcc 4. (#150086)- Rebuild 6.8.2-5 as 6.8.2-1.EL.5 for RHEL4- Conditionalize the gcc4 patch for FC4 only for now. - Disable FORTIFY_SOURCE on RHEL4/FC3 builds as it fails for some odd reason, yet it builds ok on FC4. - Added xorg-x11-6.8.2-ati-radeon-disable-broken-renderaccel-by-default.patch to disable render acceleration by default in the "radeon" driver, as it is currently broken (#136065,136941,149907,143401,143234)- Fix somehow empty patch file- Add patch for gcc 4 build problem (#150042)- Rebuild xorg-x11 with gcc 4 into FC4- Update main tarball to X.Org X11 6.8.2- [SECURITY] Added with_fortify_source option macro to spec file, to allow xorg-x11 to take partial advantage of the new gcc FORTIFY_SOURCE feature available in newer gcc releases. This option is automatically enabled in FC4 builds by rpm by default via RPM_OPT_FLAGS, however while gcc in FC3 and RHEL4 has this support also, it must be manually enabled in the spec file for it to get used. "-D_FORTIFY_SOURCE=2" gets passed to gcc when this option is enabled.- Update main tarball to X.Org X11 6.8.1.904 (6.8.2 release candidate 4) via X.Org official tarball: xorg-x11-6.8.1.904.tar.gz - Removed xorg-x11-6.8.1.903-nls-pt_BR.UTF-8-fix.patch, as it is integrated into 6.8.1.904 already.- Added new with_alternate_projectroot macro, which will eventually allow the parallel installation of multiple versions of xorg-x11 in rpm format without conflicting with each other (theoretically). The feature is intended only for Red Hat X.Org developer usage, and will not be an officially supported installation method. It is an incomplete experimental work in progress, and should not be used by mortals. THIS WILL FRY YOUR SYSTEM, DO NOT USE IT. :o)- Added patch to fix Composite gravity issue, recently added to xorg HEAD. (fdo#2230).- Added xorg-x11-6.8.1.903-nls-pt_BR.UTF-8-fix.patch from kem to fix broken pt_BR.UTF-8 support recently added to xorg cvs (fdo#2400)- Update main tarball to X.Org X11 6.8.1.903 (6.8.2 release candidate 3) via X.Org official tarball: xorg-x11-6.8.1.903.tar.gz - Updated xorg-x11.6.8.1.903-AMD64-override-default-driverlist.patch as the "i810" is now included upstream by default, but not "vmware" for AMD64 - Removed the following patches, as they are integrated into CVS now: xorg-x11-6.8.1.902-xf86pcibus-PCIX-bar-64bit-fix.patch - Added {_x11localedir}/pt_BR.UTF-8/* to file list to pick up new i18n files- Renamed build_maintainer macro to build_mharris as it is actually a personal macro and there are actually several developers maintaining the rpm for quite some time now. In case anyone is curious, I use it to tweak the spec file on occasion to build on a customized OS install that does not quite match any of our supported OS configurations. ;o) - Added a %check section to the spec file, to put post build/install sanity checks in place. - Added new post-build sanity check "with_undef_sym_test", did a local test build to confirm it works correctly, fixed a few things, then disabled it by default because Xorg needs a lot of love in order for this to be default.- Added xorg-x11-6.8.1.902-lnxLib-tmpl-sharedlibreq-fixes.patch to fix many missing library dependancies in lnxLib.tmpl, detected by examining the build log of a "with_bail_on_undef_syms" build. - Disabled with_bail_on_undef_syms, after determining that there is a bit of work that needs to be done upstream first in order for us to expect X to fully build with these linker options.- Removed dead unused i810 update patch. - Enabled with_bail_on_undef_syms, to do a test build.- Renamed the xorg-x11-6.8.0-AMD64-enable-i810-driver.patch patch to xorg-x11-6.8.0-AMD64-override-default-driverlist.patch and added the "vmware" driver to the default list of drivers to be build on AMD64. (#145588)- Added xorg-x11-6.8.1.902-xf86pcibus-PCIX-bar-64bit-fix.patch to attempt to fix bugs (#140601, 143910)- Added xorg-x11-6.8.1.902-ia64-hp-zx2-support-bug-119364.patch to add support for new Hewlett Packard IA64 ZX2 chipset. (#119364)- Update main tarball to X.Org X11 6.8.1.902 (6.8.2 release candidate 2) via X.Org official tarball: xorg-x11-6.8.1.902.tar.gz- Update main tarball to X.Org X11 6.8.1.901 (6.8.2 release candidate 1) via CVS export of CVS tag XORG-6_8_1_901 - Removed the following patches, as they are integrated into CVS now: xorg-x11-6.8.1-ati-radeon-segv-130888.patch xorg-x11-6.8.1-ati-radeon-panel-timing-param-init-130888.patch xorg-x11-6.8.0-ia64-460gx-pci-typo.patch xorg-x11-6.8.1-ati-radeon-mobility-display-detection-fix2.patch xorg-x11-6.8.1-ati-radeon-dynamic-clocks-fix.patch xorg-x11-6.8.1-libX11-stack-overflow.patch xorg-x11-6.8.1-vidmode-change-verbosity.patch xorg-x11-6.8.1-disable-dri-option.patch xorg-x11-6.8.1-xmodmap-overflows.patch xorg-x11-6.8.1-xpm-security-fixes-CAN-2004-0914.patch xorg-x11-6.8.1-xpm-security-fixes-CAN-2004-0914-sec8-ammendment.patch xorg-x11-6.8.1-r128-logout-deadlock.patch xorg-x11-6.8.1-add-missing-lucidatypewriter-font.patch - Disabled xorg-x11-6.8.1-i810-update.patch, but left it present, as only some hunks fail to apply. This leads me to believe perhaps 6.8.2 has only part of the i810 CVS head code. We need to explore this a bit before dropping the patch altogether.- Tag CVS head of latest changes, to update to 6.8.2rc1 (6.8.1.901) for the next commit- Rename xorg-x11.conf to xorg-x11-i386.conf so it works on multilib systems.- Bump version.- Drop xorg-libICE-remove-bogus-delay.patch as this was resolved in 6.8.0 (https://bugs.freedesktop.org/show_bug.cgi?id=306) - Drop XFree86-4.3.0-Xserver-xf86PciInfo-updates.patch, most of these - XFree86-4.3.0-redhat-xlib-linux-fix-avoiding-substance-abuse-job.patch is no longer needed since the unwanted code is only enabled when __sparcv9 is defined. - Remove XFree86-4.2.0-xtermresources.patch from cvs. - Use /etc/ld.so.conf.d/xorg-x11.conf instead of editing /etc/ld.so.conf in %post (#124077). - Remove XFree86-4.3.0-Xserver-dix-xkb-key-repeating-bug-CVS-backport.patch since the problem is now fixed upstream. - Remove XFree86-4.3.0-xcursorgen-check-malloc-return.patch since it would only make xcursorgen fail differently in case of OOM.- Better patch to Disable DRI on Radeon, workaround for bug #138348 .- Add disabled-by-default "dri" option to radeon driver.- Added "Obsoletes XFree86-ISO8859-2-Type1-fonts, fonts-ISO8859-2-Type1" to base-fonts subpackage, so that these ancient buggy font rpms get forcibly removed on OS upgrades from older OS releases that contained them. We removed these fonts because they are very buggy among other reasons, however they do not get removed on upgrade without flagging them Obsoleted, which causes some problems when people use the buggy fonts in applications (#129523)- Remove XFree86-4.3.0-Xserver-dix-xkb-key-repeating-bug-CVS-backport.patch since the problem is now fixed upstream. - Remove XFree86-4.3.0-xcursorgen-check-malloc-return.patch since it would only make xcursorgen fail differently in case of OOM.- Update xorg-x11-6.8.1-battle-libc-wrapper.patch to not break when some module also includes stdio.h.- Build 6.8.1-21 using build_fc3 target as 6.8.1-12.FC3.21 for FC3- Added xorg-x11-6.8.1-ati-radeon-dynamic-clocks-fix.patch which now replaces xorg-x11-6.8.1-ati-radeon-7000-smp-lockup.patch as a fix for a lockup problem that occurs on some models of Radeon on SMP systems, related to the dynamic clocks code. Fixes bugs: (#138108, 138348, 133526, 138539, 138779, fdo#1912) - Merge to FC-3 branch and tag as 6.8.1-12.FC3.21, and to RHEL-4 branch and tag as 6.8.1-21.EL- Disabled with_fonts again - Added xorg-x11-6.8.1-xpm-security-fixes-CAN-2004-0914-sec8-ammendment.patch to fix bugs introduced into libXpm in the previous security patch for CAN-2004-0914 which cause gimp to fail to save Xpm files (#140815, 141047, fdo#1924)- Update xorg-x11-6.8.1-ati-radeon-7000-smp-lockup.patch to the new patch from Alex Deucher , provided in https://bugs.freedesktop.org/show_bug.cgi?id=1912.- Enabled with_fonts for a single build, to update fonts-xorg package- Updated a few patches to apply with -p0 for consistency - Reordered some patches by patch number - Change libI810XvMC inclusion to be ifarch ix86/ia64/x86_64 only, since we know Intel video exists on those arches. This is better than excluding it on arches we know it is not available on, because the list is shorter this way, and exact. Less maintenance this way.- Remove libI810XvMC from PPC, as there is no Intel video hardware available for PPC architecture and this is an Intel video specific client side lib. - Added xorg-x11-6.8.1-r128-logout-deadlock.patch to fix lockup on logout bug on Rage128 with DRI (#138822) - Added xorg-x11-6.8.1-add-missing-lucidatypewriter-font.patch from upstream to fix bug with fonts that got inadvertently excluded from the X11 6.8.1 release. Our fonts-xorg package is being updated to include these additional fonts now as well (#139108, 139121, fdo#1560) - Enable xorg-x11-6.8.1-i810-update.patch for all OS version builds now.- Add libI810XvMC back on PPC filelists. - Move xorg-x11-6.8.1-battle-libc-wrapper.patch into fc4_build section.- Add xorg-x11-6.8.1-battle-libc-wrapper.patch to work around build error that pops up when system headers #define printf.- Added xorg-x11-6.8.1-ati-radeon-7000-smp-lockup.patch to fix the hard lockup happening on SMP boxes with Radeon 7000 cards (#138108).- Added xorg-x11-6.8.1-nls-indic-locales.patch to add support for some indic locales. BLOCKER (#133880) - Minor fix in application of vidmode verbosity patch - Added build_fc4 and build_rhel4 spec file build targets - Enabled DRI for PPC only on FC2/FC3/FC4 builds for Fedora/PPC users. - Conditionalized i810-update patch to only apply for build_fc4 as it is currently experimental.- Add xorg-x11-6.8.1-i810-update.patch which contain fixes for i810 chipsets backported from CVS HEAD as of November 16th (#132267).- Added xorg-x11-6.8.1-xpm-security-fixes-CAN-2004-0914.patch to fix a number of Xpm issues found by Thomas Biege (#136169)- Remove xorg-x11-6.8.1-ati-radeon-mobility-display-detection-fix.patch since it has been replaced with the -fix2 patch, remove XFree86-4.3.0-redhat-version-change-detection.patch since it doesn't work for Xorg. - Remove XFree86-4.3.0-remove-copyright-symbol-to-fix-UTF-8.patch since it is now fixed upstream.- Drop XFree86-4.3.0-Xserver-common-fix-broken-MIN-macro-definition.patch since it's harmless and it's fixed in Xorg HEAD now. - Removed commented out xbiff, xeyes and xditview files from file lists, they are in CVS if we need them. - Updated xorg-x11-tools description so it doesn't advertise stuff it doesn't contain (xbiff and xeyes). - Drop XFree86-4.3.0-xbiff-file-heirarchy-standard.patch as we don't ship xbiff. - Drop XFree86-4.2.0-xtermresources.patch as we don't ship in tree xterm anymore. - Drop XFree86-4.3.0-redhat-lock-version.patch as Xorg don't print scary warnings for betas.- Edit xorg-x11-6.8.1-xmodmap-overflows.patch to fix buffer overflow crash (#138743).- Add NULL check in xorg-x11-6.8.1-init-origins-fix.patch to prevent crashes on single head setups.- Added xorg-x11-6.8.1-ati-radeon-mobility-display-detection-fix2.patch to fix ATI Radeon Mobility display detection issue. (#132865, 133147, and upstream bugs: (fdo #1599, 1301, 1303, 1374, 1482, 1514, 1622, and 1631, and numerous others not on our lists))- Set with_fonts based off build_fc2 macro, so it is enabled for FC2 builds but remains disabled for FC3 and other builds. Makes it easier to rebuild for FC2/FC1. - Fix to make Vera font really get deleted this time for real. ;o)- Add xorg-x11-6.8.1-init-origins-fix.patch, which fixes a crash for some invalid xorg.conf combinations (#134967).- Do not include the Bitstream Vera fonts when doing builds with with_fonts enabled, as we ship Vera in separate packaging already. - When doing with_fonts builds to generate the fonts-xorg tarball, remove various Syriac fonts which contain bad codepoints: SyrCOMCtesiphon.otf SyrCOMKharput.otf, SyrCOMMalankara.otf, SyrCOMMidyat.otf, SyrCOMQenNeshrin.otf SyrCOMTurAbdin.otf, SyrCOMUrhoyBold.otf, SyrCOMUrhoy.otf (bug #97951)- Added xorg-x11-6.8.1-ati-radeon-segv-130888.patch from Kevin Martin to fix linked list related SEGV for bug (#130888) - Added xorg-x11-6.8.1-ati-radeon-panel-timing-param-init-130888.patch by Kevin Martin to fix panel timing issue in Radeon driver (#130888)- Add xorg-x11-6.8.1-xmodmap-overflows.patch (#83720, #103161).- Add xorg-x11-6.8.1-disable-dri-option.patch, which adds XF86Free-DRI to the list of extensions that can be disabled using the -extension command line switch (#135071).- Added xorg-x11-6.8.1-vidmode-change-verbosity.patch which changes the default verbosity level of Xvidmode extension debugging log messages in the X server from "1" to "> DEFAULT_XF86VIDMODE_VERBOSITY" which is set to 3. This fixes a power consumption problem on laptops running on battery power, where when a screensaver kicks in and calls xf86GetVidMode(), the X server generates excessive log messages which cause the hard disk to spin up if it was in powersave mode, thus reducing the battery life. Since the currently logged messages are pretty useless generally speaking, this fixes the problem by increasing the verbosity level required before the messages end up in the log file. (#128305)- Added post/postun scripts to the deprecated-libs subpackage, copied versions of those in the libs subpackage (#134424)- Implemented xorg-x11-6.8.1-libX11-stack-overflow.patch to fix stack overflow in libX11 which was discovered by new security features of gcc4 (#132885)- Bump to 6.8.1 and rebuild to quell beehive complaining - Disable "parallel_build", as it is still broken- Update main tarball to the final X.Org X11 6.8.1 release via CVS export of CVS tag XORG-6_8_1 - this obsoletes previous 6.8.1 tarball which was a upstream release mixup. - Added xorg-x11-6.8.0-libXdmcp-build-with-PIC.patch to make libXdmcp.a build with PIC flags to fix (#131130) - Fixed changelog dates and made changelog version-release strings consistent - Re-enable "parallel_build 1" to test if it works currently, so we can take advantage of SMP machines better- Update to xorg 6.8.1 - Change Source: to use xorg-6.8.2.tar.gz since that is what the tarball from http://freedesktop.org/~xorg/X11R6.8.1/src/single/ is called.- Enable "#define HasDevRandom YES" in host-$arch.def for bug (#126205) - Remove "Requires: xorg-x11, xorg-x11-libs = 6.8.2-1.EL.13.16" from twm subpackage, as the libs twm is linked to should automatically be picked up by rpm "find-provides", and xterm moved to a separate package a long time ago, so both of these should no longer be needed. (#123423) - Added xorg-x11-6.8.0-ia64-460gx-pci-typo.patch to fix typo in Intel 460gx PCI handling code on ia64. (#126586) - Remove "XtermWithI18N" from host.def as we build xterm externally - Added SElinux fix to xfs initscript (#130421,130969)- buildrequire redhat-rpm-config, use brp-compress from that (fixes file conflicts)- Fix typo in test for xorg.conf (#127711).- Add with_wacom_driver option and set it to 0 (#132438, #132440).- Update main tarball to the final X.Org X11 6.8.0 release via CVS export of CVS tag XORG-6_8_0 - Add "BuildXaw NO" to host-$arch.def to disable building of libXaw 8 library - Add "XtransFailSoft YES" to host-$arch.def to fix (#129622) - Moved "archexec" script from main xorg-x11 subpackage to xorg-x11-devel subpackage since it is only used by things in the devel subpackage and we do not want to have the devel subpackage have a forced dependancy on the main package. (#132121) - Added "Conflicts: xorg-x11 < 6.8.0-1" to xorg-x11-devel for archexec subpackage move, so upgrades work correctly - Removed libXaw.a from file list, as 6.8.0 no longer builds static libXaw 7 even when libXaw 8 is disabled by "BuildXaw NO" - Rework the _x11localedir file lists to fix rpm build problems- Remove various Syriac fonts which contain bad codepoints: SyrCOMCtesiphon.otf SyrCOMKharput.otf, SyrCOMMalankara.otf, SyrCOMMidyat.otf, SyrCOMQenNeshrin.otf SyrCOMTurAbdin.otf, SyrCOMUrhoyBold.otf, SyrCOMUrhoy.otf (bug #97951)- Remove hard coded dependancy on Glide3 from xorg-x11 package, as the tdfx DRI driver has used dlopen() with Glide3 since XFree86 4.2.1 or so, and no longer has a build time or runtime dependancy on Glide3 being present. If Glide3 is present, DRI will work (theoretically), but if it is not present, the tdfx driver should still function correctly, so we can get rid of this dependancy now, as we are removing Glide3 from some of our OS products.- bump release again- Bump release field and rebuild unmodified to handle an off-by-one error in release field of font-utils subpackage dependancy- Enable "#define HasDevRandom YES" in host-$arch.def for bug (#126205) - Remove driver list overrides for all architectures except ia64, as ppc was missing the "nv" driver. In the future, we will use patches to the Imake config directory to disable individual drivers on particular architectures. - Disable building of fonts by setting with_fonts 0, as we now provide the fonts from a separate src.rpm which builds noarch packages, so that we no longer regenerate fonts every single X build, and people do not have to redownload fonts every single xorg-x11 update when the fonts never change anyway. - Added xorg-x11-6.8.0-Imake-add-BuildFontDevelUtils-macro.patch to allow building and inclusion of ucs2any and bdftruncate when doing "with_fonts 0" builds, as "BuildFonts NO" causes these utils to be excluded. - Change main package dependancy from xorg-x11-base-fonts to base-fonts which is a virtual provide in both the xorg-x11-base-fonts and fonts-xorg-base packages now, to maximize install/upgrade compatibility. - Moved "ucs2any" from the xorg-x11-tools subpackage to the font-utils subpackage where it belongs, and added a Conflicts line to the font-utils package to handle upgrades. - Moved "fonts/util" subdir from xorg-x11-base-fonts, to font-utils package, as the utilities that use it reside there, and it never made sense in the base-fonts package anyway really. Added Conflicts line to font-utils to handle upgrades.- Remove "xtt" module line from config file on upgrades, as X.Org no longer supplies that module, since the "freetype" module now includes the functionality that was only available in the "xtt" module before.- Update main tarball to CVS export of CVS tag XORG-6_7_99_903 snapshot, a.k.a. 6.8.0 RC3 - Added "xmessage" back upon request (#131076) - Updated file list to change libXevie.so.1.0.0 to libXevie.so.1.0 due to upstream .so versioning change.- Add defattr flag to Xdmx subpackage (#130870)- Added new rpm macro 'with_libs_data' to Conditionalize the libs-data subpackage. This allows us to disable the "libs-data" subpackage now, as rpm multilib should properly handle both architectures sanely, so the workaround of having the library data in a separate package is no longer needed. By setting 'with_libs_data' to 0, it will cause the libs-data files to be included in the "libs" subpackage as it was in the past, and the "libs-data" package will get obsoleted. This will simplify the rpm packaging a slight bit for the future.- Remove a bunch of legacy Xaw/Xt applications which really should not be part of X11 sources (xbiff,xditview,xeyes,xmessage)- Update archexec with additional cleanups, and make host.def match the version of host.def backported to XFree86 4.3.0 for RHEL3-U3, to maximize upgrade compatibility.- Added "archexec" script as a generic solution to the xft-config/gccmakedep problem. - Renamed xft-config and gccmakedep to -, and added symlinks from archexec to , so these two files do not conflict when installed on multilib systems.- Install the host-.def file where host.def gets installed.- Initial attempt to make host.def architecture neutral, and split the architecture specific things onto host-.def specific files, to allow multiple xorg-devel packages to be installed on multilib systems, such as x86_64, ppc64, and s390x.- Update main tarball to CVS export of CVS tag XORG-6_7_99_902 snapshot, a.k.a. 6.8.0 RC2 - Removed patches that are now included in upstream sources: - xorg-x11-6.8.0-glx-install-dri-modules-in-correct-place.patch- Set "BuildXprint NO" "BuildXprintClients NO" "BuildXprintLib YES" so we get *only* the Xprint libraries for backward compatibility with existing dynamically linked applications. - Added new "xorg-x11-deprecated-libs", and moved the Xprint runtimes to it for backward compatibility for now, until we remove libXp entirely at some point in the future. - Replaced remaining spec file package description references of "XFree86" to something X11 implementation neutral. - Exclude the two new Xprint libraries libXprintAppUtil.a, libXprintUtil.a as we do not ship anything that links to them, so do not need them for back compat.- Update main tarball to CVS export of CVS tag XORG-6_7_99_901 snapshot, a.k.a. 6.8.0 RC1 - Removed patches that are now included in upstream sources: - xorg-x11-6.7.0-mga-storm-sync-fix.patch - xorg-x11-6.8.0-ppc64-XorgServer-not-XF86Server-fix.patch - Update xorg-x11-6.8.0-redhat-custom-startup.patch due to Imakefile change - Disabled xorg-6.8.0-redhat-libGL-exec-shield-fixes.patch for now, as Xorg changes have broken it again and I do not want to port it forward 10 times. Once 6.8.0 is finalized, we will port it forward to Mesa CVS and get it into the Mesa CVS head *first*, then we will backport it to 6.8.0. This way we do not have to maintain and port this patch once a week. - Removed unwanted xprint file deletion from specfile as "BuildXprint NO" actually works now. - Added Xevie libraries to file lists.- Remove unneeded pre-build of ucs2any.c as it is now part of xorg stock - Various minor cleanups to specfile - Remove 'restest' Xresource extension sample client, as we now have 'xrestop' which is way better. - Update xorg-x11-6.8.0-glx-install-dri-modules-in-correct-place.patch to use MODULEDIR instead of BUILDMODULEDIR (fdo #1057)- Nuke libXp.a as it is unneeded (libXp got enabled for libXaw)- Added xorg-6.8.0-glx-install-dri-modules-in-correct-place.patch to correct the location where dri modules get installed on lib64 arches (#129668) - Enable libXp.so.* temporarily as libXaw is cancerously attached to it- Added code to the rpm pre script to munge the config file to change the "keyboard" driver to "kbd" driver on upgrades, as XOrg no longer supports the old "keyboard" driver. - Use perl -p -i -e in pre rpm script for all config file munging instead of grep and temporary files.- Update main tarball to CVS export of CVS tag XORG-6_7_99_2 snapshot - Removed patches that are now included in upstream sources: - xorg-x11-6.8.0-ppc64-support-updates.patch - xorg-x11-6.8.0-fix-xprint-install.patch - xorg-x11-6.8.0-build-fix-for-non-dri-builds.patch - Remove /etc/init.d/xprint, as we do not ship/support Xprint - Set "#define BuildXprint NO" in host.def - Add xorg-x11-6.8.0-fix-BuildXprint.patch to fix "BuildXprint NO" - Add xorg-x11-6.8.0-ppc64-XorgServer-not-XF86Server-fix.patch to fix ppc64 build issue due to upstream renaming of XF86Server to XorgServer - Added "XorgServer NO" to host.def for non-server builds- Update main tarball to CVS export of CVS tag XORG-6_7_99_1 snapshot - Removed patches that are now included in upstream sources: - xorg-x11-6.7.0-xkb-sysreq.patch - xorg-x11-6.7.0-mga-enable-video-rom-before-using.patch - xorg-x11-6.7.0-xset-man-update.patch - xorg-libX11-zh_CN.UTF8-crash-fix.patch - xorg-x11-6.8.0-i810-remove-stale-sdk-header.patch - Update xorg-x11-6.8.0-redhat-custom-startup.patch due to Imakefile change - Fix glitch in macro that disables DRI on ppc builds for fc3/rhel4 - Implemented xorg-x11-6.8.0-build-fix-for-non-dri-builds.patch patch to fix a compile time bug in the Radeon driver render acceleration on non-DRI builds, where the driver tests if CPStarted is set before calling RADEONInit3DEngineForRender(), however CPStarted is not available in non-DRI builds. Also, RADEONInit3DEngineForRender() already correctly handles both the DRI and non-DRI cases both at compile and runtime, and checks if CPStarted is set internally. (xorg #1033)- Delete various X server driver manpages on ppc64 s390, s390x which are showing up now for some odd reason, even when servers are not being built. - Do not include libI810XvMC on ppc/ppc64/s390/s390x filelists- Updated to X.Org X11 CVS development and branched specfile. Chose arbitrary version number of "6.7.99.0" as XOrg confusingly calls development builds "6.7.0" still, and we need differentiation for sanity. Kevin mentioned he would tag 6.7.99.1 in CVS once the final version 6.8.0 is approved by the XOrg BOD, so everything should work out ok version-wise. - Removed patches already included in new upstream snapshot: - xorg-x11-6.7.0-ati-driver-message-cleanups.patch - xorg-x11-6.7.0-ati-radeon-pagesize.patch - xorg-x11-6.7.0-ati-radeon-7000m-dell-server.patch - xorg-x11-6.7.0-siliconmotion-fboffset-fix-from-cvs-head.patch - xorg-double-cast.patch - XFree86-4.3.0-build-libXinerama-before-libGL-for-via-driver.patch - xorg-use-linux-pci-only.patch - xorg-x11-6.7.0-allow-XF86ExtraCardDrivers-on-AMD64.patch - xorg-r200-uninitialized-variable-used.patch - Disabled ati-radeon-forcelegacycrt and xorg-x11-6.7.0-xterm-make-optional patches, as they need porting forward and are very low priority - Removed voodoo driver tarballs, as a newer version of the driver is now included in stock Xorg - Update xorg-x11-6.8.0-Xserver-increase-max-pci-devices.patch for 6.8.0, as the upstream default changed from 64 to 128, but we want it to be 256. - Update xorg-x11-6.8.0-AMD64-enable-i810-driver.patch for 6.8.0, as upstream currently still does not enable the i810 driver on AMD64 arch for some reason. - Update xorg-x11-6.7.0-ppc64-support-updates.patch for 6.8.0 - Updated xorg-6.8.0-redhat-libGL-exec-shield-fixes.patch for 6.8.0 - Commented out the codeconv dir - no longer present - Commented libXfontcache out of file lists, no longer present - Added new libraries to file lists: libXcomposite, libXdamage, libXfixes, - Added pkg-config files xcomposite.pc, xdamage.pc, xfixes.pc, xrender.pc - Added all of the new Xprint junk to the xprint exclusion list - Created new xorg-x11-Xdmx subpackage to contain all of the DMX related files required to run DMX. - Removed with_via_driver option as it is no longer needed - Removed "WORLDOPTS=" from make World invocation as upstream defaults to this now for some time. - Added xorg-x11-6.8.0-i810-remove-stale-sdk-header.patch as workaround for header file inclusion problem (Kristian Hogsberg, fdo-bz #995) - Added xorg-x11-6.8.0-fix-xprint-install.patch (Kristian Hogsberg)- Disabled XFree86-4.3.0-ia64-new-elfloader-relocations.patch as it fails the build on ia64.- Disabled XFree86-4.2.99.901-chips-default-to-noaccel-on-69000.patch and XFree86-4.2.99.901-chips-default-to-swcursor-on-65550.patch for the time being at least in order to see if anyone still reports the problem, in which case we can narrow down the bad primative, report upstream with workaround patch that disables accel. (#82438) - Removed XFree86-4.0.1-nohardware.patch as it is ancient, undocumented, and probably not needed anymore. - Removed unnecessary or no longer needed patches, including: - XFree86-4.3.0-ati-r128-chip-names-touchup.patch - XFree86-4.3.0-nls-cjk-utf8.patch, XFree86-4.0-xdm-servonly.patch - XFree86-4.2.99.4-x86_64-glx-nopic.patch - XFree86-4.3.0-xterm-can-2003-0063.patch - XFree86-4.3.0-SDK-add-missing-includes-for-synaptics.patch- Updated and renamed following patches from XFree86-* to xorg-x11-* while cleaning up and documenting them for submission upstream: - xorg-x11-6.7.0-xset-man-update.patch - xorg-x11-6.7.0-xkb-sysreq.patch - xorg-x11-6.7.0-ati-driver-message-cleanups.patch - xorg-x11-6.7.0-ati-radeon-pagesize.patch - xorg-x11-6.7.0-ati-radeon-forcelegacycrt.patch - Removed patches that are tested and confirmed to be no longer needed: - xorg-x11-6.7.0-cpp-s390x.patch - XFree86-4.2.99.3-locale-alias-gb18030.patch - XFree86-4.3.0-nls-compose-en_US.UTF8-lxo.patch - XFree86-4.3.0-ati-ia64-no-nonpci-ioport-access.patch - XFree86-4.3.0-ia64-slowbcopy.patch - XFree86-4.3.0-ia64-slowbcopy2.patch - Disabled the sis DRI driver using with_sis_dri macro, as it is currently too experimental (#126821) - Added xorg-x11-6.7.0-siliconmotion-fboffset-fix-from-cvs-head.patch taken from X.Org CVS head to try and fix bug (#127278)- Test build with xorg-x11-6.7.0-cpp-s390x.patch disabled- Updated and renamed following patches from XFree86-* to xorg-x11-* while cleaning up and documenting them for submission upstream: - xorg-x11-6.7.0-mga-enable-video-rom-before-using.patch - Removed patches that are tested and confirmed to be no longer needed: - XFree86-4.3.0-r128-unresolved-symbols.patch - XFree86-4.3.0-ati-r128-Xv-ecp-divisor.patch - XFree86-4.3.0-gb18030-20020207.patch - XFree86-4.3.0-gb18030-xtt.patch, XFree86-4.3.0-gb18030-enc.patch - XFree86-4.3.0-nv-riva-dualhead-fixes.patch - XFree86-4.3.0-siliconmotion-Xv-stability-fix.patch - XFree86-4.2.99.901-007_fix_xfree86_man_version_string.patch - XFree86-4.2.99.901-008_doc_extensions_fix.patch - XFree86-4.3.0-Xserver-includes-ansify-prototypes.patch - XFree86-4.3.0-missing-SharedXfooReqs.patch - XFree86-4.3.0-libXrandr-missing-sharedreqs.patch - XFree86-4.3.0-libXpm-missing-sharedreqs.patch - XFree86-4.3.0-missing-lib-sharedreqs.patch - XFree86-4.3.0-Mesa-SSE-fixes-from-MesaCVS-v2.patch- Remove temporary config file left over after upgrade (#127187) - Add libGL-devel and libGLU-devel virtual provides to xorg-x11-devel subpackage. Both are provided for the major library version (1 in both cases), and for the major.minor (1.2 for libGL, 1.3 for libGLU). This is being added to allow 3rd party packagers to add BuildRequires in an agnostic manner for apps that use libGL/libGLU, which allows any libGL/libGLU implementation devel package to be installed and meet the dependancy, as long as it also provides the same virtual provide mechanism.- Added xorg-x11-6.7.0-AMD64-enable-i810-driver.patch to patch the i810 driver into the default driver list instead (#126687)- Enabled XF86ExtraCardDrivers on x86_64, and added i810 driver to it, to implement feature request from Intel (#126687) - Added xorg-x11-6.7.0-Xserver-increase-max-pci-devices.patch to Increase the maximum number of PCI devices the X server scans, by changing the compiled in constant MAX_PCI_DEVICES from 64 to 256, as a lot of newer x86, ia64, ppc, ppc64, AMD64 hardware exists which may have more than 64 devices. (#126164)- Fixed bug in mga driver which caused hangs on some Matrox Mystique boards of revision 0->2, which were caused by a previous upstream bugfix for another issue. xorg-x11-6.7.0-mga-storm-sync-fix.patch (#124028) - Added xorg-x11-6.7.0-ati-radeon-7000m-dell-server.patch to add support for custom ATI hardware made for Dell. (#122190)- rebuilt- Added rpm script munging of X server config file back to spec file, which got inadvertently left out of the 6.7.0-2 build. This fixes #120858 and its bretheren.- Add -Wa,--noexecstack to AsCmd, in order to force all assembler files to use GNU stacks. This was accomplished in an alternative way in our XFree86 4.3.0 packaging via XFree86-4.3.0-redhat-exec-shield-GNU-stack.patch which patched every assembler file to add a .note.GNU-stack section, which is considered a superior solution, but our patch was non-portable to non-GNU systems. This is a cleaner hack for now. (FC2BLOCKER #122708) - Modified main rpm pre script to massage the X server config file(s) to remove XkbRules lines to help avoid (FC2BLOCKER #120858 and a zillion duplicates), and to ensure that the only active X server config file is xorg.conf (if any is present), and any other XFree86 4.x config files are renamed to .obsolete, which will help to minimize config file name confusion, and provide some sanity on upgrades, as all users will use xorg.conf by default regardless now. (FC2BLOCKER #122454) (Fix based on patch from wtogami)- Added xorg-x11-6.7.0-libxf86config-monitor-freq-fix.patch to fix a problem caused by gratuitous upstream libxf86config changes which force the HorizSync and VertRefresh lines to always be written out to the config file in commented out form. (FC2BLOCKER #120950, 122341, 122573, 122439, 122072, 122461, 121946, 121717, others not yet duped)- add xorg-use-linux-pci-only.patch, fixes bugs #118130 & #120520- Further package description cleanups to remove more references to "XFree86" and replace them with more neutral wording, and also fix (#119670)- Fix pre script for loop error created by accidentally committing half worked code.- Updated spec file package descriptions to remove references to "XFree86" and update them to "X Window System" or "Xorg X11" as appropriate. - Updated the following patches to work with xorg.cf instead of xfree86.cf: - xorg-x11-0.6.6-allow-XF86ExtraCardDrivers-on-AMD64.patch - xorg-x11-0.6.6-fix-BuildXFree86ConfigTools.patch - xorg-x11-6.7.0-xterm-make-optional.patch - xorg-x11-6.7.0-ppc64-support-updates.patch- Updated file lists to handle upstream renaming of X server binary from "XFree86" to "Xorg", server manpage, and getconfig config file. - Added new ghost files for new server config file /etc/X11/xorg.conf - Removed legacy compatibility symlinks for the old XF86Config file from /usr/X11R6/lib/X11 as X.org X11 uses /etc/X11/xorg.conf and there is no legacy to be compatible with. - Other specfile cleanups to use /etc/X11/xorg.conf as the config file, and flag all places of XF86Config files with FIXME warnings of deprecation- Update main tarball to official X11 Release 6.7.0 from CVS export of the upstream release tag "xo-6_7_0" off of the "XORG-RELEASE-1" branch of CVS - Removed xorg-redhat-ia64-plt-prot-exec-fix.patch as it is merged in 6.7.0- Removed build_psyche build target, renamed build_rawhide to build_fc2, and added new build_maintainer custom build target - Various spec file cleanups, renamed macros to foo_bar format, etc.- Updated our ppc64 support patch and submitted upstream to fix (FDO #303) - Renamed xorg-ppc64-support-updates.patch to xorg-x11-ppc64-support-updates.patch to establish consistency with the names my xdiff utility produces. - Added xorg-x11-0.6.6-allow-XF86ExtraCardDrivers-on-AMD64.patch to allow add on drivers to be specified using XF86ExtraCardDrivers in host.def on AMD64 architecture, in order to enable the "voodoo" driver. - Enable the "voodoo" driver for AMD64 in addition to x86. - Removed build_shrike, build_yarrow, and build_taroon build target macros and usage, as the package is not designed or intended to be used on those OS releases, and there are various potential problems that could arise. - Updated freetype and freetype-devel dependancies to require version 2.1.7, as that is what version ships in the upstream sources, and there have been some font display related bugs reported by people who have recompiled the SRPM on older OS releases, and we do not support that anyway. - Fixed bugs in spec file command substitution '[' tests discovered while updating the build target macros. - Disabled BuildRequires: Glide3, Glide3-devel dependancies as Glide is opened via dlopen() for a while now, so it should not be needed at compile time any more.- Updated xorg-ppc64-support-updates.patch with fix for endian related issue (FC2 BLOCKER #119045) - Removed obsolete_xfree86 conditional and usage, we hard code the Obsoletes lines now. - Restored "Conflicts" lines that I had previously removed in the transition from XFree86 to Xorg X11, as they seem to still be needed on upgrades, or it is possible there will be upgrade failure issues, in particular when upgrading from Red Hat Linux 7.x or older releases to FC2. - Added xorg-libX11-zh_CN.UTF8-crash-fix.patch to fix (FC2 BLOCKER #119032)- Fixed a bug in xorg-redhat-libGL-exec-shield-fixes.patch (FC2 BLOCKER #119818) - John further updated above patch, and tagged the .3 release.- Added new open source Nvidia DRI 3D driver today, and enabled it on x86, ia64, and x86_64 only, as those are the only architectures that have been tested. It builds cleanly with xorg, but is totally untested so far.- Added xorg-ppc64-support-updates.patch back, as PPC64 fixes were reverted upstream .- Updated main xorg tarball to CVS snapshot 2004_03_30 from today. - Removed XFree86-4.3.0-radeon-dpms-on-dvi-v2.patch as it should no longer be needed with current Radeon driver. - Removed patches already merged into new upstream tarball, including: - xorg-redhat-elfloader-linux-non-exec-stack.patch - xorg-x11-addrinuse.patch - xorg-x11-Xft-freetype-bitmap-font-fix.patch - Split out xorg-redhat-ia64-plt-prot-exec-fix.patch from libGL-exec-shield patch, as it was unrelated to libGL-exec-shield fixes. (#119324) - Removed fonttosfnt* from the file lists as it has been removed upstream now - Updated file list for libXft.so.2.1.2- Added xorg-r200-uninitialized-variable-used.patch to CVS, as my last commit neglected to include the file.- reenable libGL exec shield patch - also fix bug 119324 ia64 exec permissions on PLT in elfloader- Fixed uninitialized variable access in Radeon R200 driver, in r200_pixel.c with xorg-r200-uninitialized-variable-used.patch FC2t2 TARGET (#116661)- Really added the xorg-x11-Xft-freetype-bitmap-font-fix.patch patch this time as it was inadvertently left out of 0.0.6.6-0.0.2004_03_11.8 by mistake- Added xorg-x11-Xft-freetype-bitmap-font-fix.patch from Xft 2.1.6 to Xorg included Xft to fix - FC2t2 BLOCKER (#118587,118822) - Remove the rpm pre scripts from the library subpackages as they did not end up solving the XFree86 -> xorg-x11 ld.so.conf upgrade problem. - Disable userdel/groupdel in xfs preun script, as it is harmless to leave the xfs user/group around as it is a registered account ID anyway, and there is a rather unsolveable problem on upgrades from one package providing xfs to a new one otherwise. - FC2t2 BLOCKER (#118145) - Added "triggerpostun xfs -- XFree86-xfs" script, to work around a bug in the XFree86-xfs postun script, which results in the special xfs user account being inadvertently removed, causing xfs to run as the root user, and also resulting in xfs not being activated by chkconfig. This trigger executes right after the XFree86-xfs postun script, and ensures that the xfs user exists, and that the xfs initscript is properly added by chkconfig, however this problem will recur again if we ever rename the xorg-x11 package to something else, or switch to a different X11 implementation altogether, since the xfs postun still will remove chkconfig on "rpm -e" or "rpm -U" to a different package name, but that problem is not easily solved other than by adding a new trigger every time the package name changes, unless someone has a better idea - FC2t2 BLOCKER (#118145,118818) - Added "triggerpostun libs -- XFree86-libs" script to work around a bug in the XFree86-libs postun script, which results in _x11libdir being removed from ld.so.conf when upgrading from XFree86 to xorg-x11. This trigger executes right after the XFree86-libs postun script and puts _x11libdir back in ld.so.conf - FC2t2 BLOCKER (#118448,118851,118993) - Use "/sbin/service" instead of "service" in rpm scripts. - Add "Requires(preun,postun): /sbin/service" to xfs package dependancies.- Fixed -libs subpackage postun script by commenting out the broken empty if block (#118890,118949) - Added xorg-x11-addrinuse.patch to fix bug in X.org Xtrans code which causes bind() failures to fail as if errno was set to -EADDRINUSE (#118950) - Set "LinkGLToUsrLib NO" because the symlinks Imake creates are absolute, and we want relative symlinks. A user reported this bug, so I have added code to %install section to create relative symlinks in a similar but cleaner way to our XFree86 4.3.0 packaging.- Added new rpm pre scripts for the libs, Mesa-libGL, and Mesa-libGLU packages which add _x11libdir to ld.so.conf if it is missing, and then run ldconfig, in order to work around a race condition in the XFree86-libs-4.3.0-64 and older XFree86-libs packaging which could cause _x11libdir to be removed from ld.so.conf when other packages are still using it, including xorg-x11, causing upgrades to fail (#118448) - Removed the offending code from xorg-x11-libs postun script so that this problem is not preserved in xorg-x11 packaging. Also updated our XFree86 packaging in cvs, so future XFree86 erratum for Fedora Core 1 and other OS releases will help to reduce the chances of someone getting hit by this upgrade problem. - Added initial support to spec file for using the CVS Revision of the spec file in the package Release field. Graciously stolen from the kernel spec file. This is currently unused, but may be used in the future.- Added versions to some of the XFree86 compabitility virtual provides, which include "Provides: XFree86-devel = 4.4.0", "Provides: XFree86-font-utils = 4.4.0", "Provides: XFree86-libs = 4.4.0" - Added virtual provides of xdm, base-fonts, Xnest, xauth, libGL, libGLU, to each relevant subpackage. They are all intentionally unversioned, so that things that require xdm/Xnest/etc. in a version/implementation agnostic manner, should use "Requires: xdm" instead of requiring a specific implementation or version of xdm. If some package really does require a specific implementation or version of xdm, then it should still use the full non-virtual package name, as it is going to be both implementation and version specific anyway, so versioning these particular virtual provides does not make sense.- Removed a bunch of the virtual Provides I added previously as they create problems with previous usage of Conflicts lines in some subpackages, and likely with other software, since they were not versioned Provides. The other option would be versioning the provides, but I am not convinced that is the right approach. Instead, I have removed them and will add things for compatibility only as problems arise. - Added virtual "Provides: foo" to the Xvfb, xfs, twm, and font-utils subpackages, where "foo" is the subpackage name with the "XFree86-" part stripped off so that apps can use for example "Requires: xfs" to get an xfs font server, instead of requiring the XFree86 implementation specifically. This allows us to have X11 implementation agnostic dependancies, which makes it easier to switch X11 implementations without breaking a pile of package dependancies. - Added ld_so_conf_add_x11libdir macro, to clean up the rpm scripts a bit- Added Obsoletes: tags to each subpackage in order to be able to properly replace XFree86 on upgrades cleanly - Added virtual Provides: to most subpackages for compatibility with XFree86 packaging on upgrades for packages that hard code specific XFree86 dependancies. Note that this may change in the future depending on the problems and needs that arise. - Remove {_libdir}/libGL.so.1.2 symlink from Mesa-libGL subpackage as unneeded - Disable libGL exec-shield patch for libGL as it has a slightly elusive glitch - Remove all old XFree86 Conflicts lines from packaging, as they should not be needed now since the package names have changed and the old ones are now fully obsoleted- Removed with_vmware conditional, and now unconditionally include vmware_drv - Removed both the with_new_xft and with_new_xrender macros and their usage - Removed version from virtual Xft Provides, as it is probably unnecessary - Ported redhat-libGL-exec-shield-fixes.patch from 4.3.0 and added it as xorg-redhat-libGL-exec-shield-fixes.patch- Various spec file cleanups, and s/XFree86/X Window System/ etc. - Use %{name} instead of hard coded XFree86 in dependancies for xdm - Remove triggerpostun -- XFree86, as that was for 3.3.6 compatibility only - Fix dri module file lists to prevent double inclusion problem - Disable TLS support for libGL and the DRI drivers until it is ported forward- Removed with_new_render hacks from specfile because new Render is included now, and upstream imake can deal with 3 level .so versions - Updated libXft/libXrender .so.* versions in rpm file lists - Updated spec file bits for moving GL includes into OpenGL ABI for Linux mandated location - Removed spec file hack to make libGL be visible in OpenGL ABI for Linux mandated location, and enabled LinkGLToUsrLib YES in host.def to accomplish the same instead. - Replaced .so version specific spec file hack to make libGLU be visible in OpenGL ABI for Linux mandated location, with a new hack that is more robust - Updated file list for libXcursor.so.1.0.2- Updated to CVS snapshot 2004_03_11 - Added missing .so symlinks to -devel subpackage file list that did not get detected by the rpm missing_files_terminate_build test, but were present in the RPM_BUILD_ROOT - Fixed bad dates for last few entries in spec file changelog - Applied the xorg-mharris-radeon-agp-detect-via-pci-cap-list.patch and libICE-remove-bogus-delay.patch to upstream CVS and removed patches - Downgraded freetype-devel dependancy to 2.1.4 again, so the package builds more easily on older OS releases for anyone brave enough to do so (ok, ok, so it is for me ok!) - Added xorg-Imake-make-icondir-configurable.patch, ported forward from our XFree86 4.3.0 patch - Explicitly list each individual shared library in the file lists instead of using globs, as this makes it easier to know when .so versions change, perhaps accidentally and catch problems sooner. Globs were used before mainly to ensure files did not get excluded accidentally, however rpm now checks for files in the buildroot that do not end up in the final packaging - Remove XFree86-4.3.0-redhat-xft-loadtarget.patch as the Xft in xorg CVS now includes this functionality already, however it is conditionalized at build time on the definition of FC_HINT_STYLE being defined, so we need to ensure this is getting defined at build time for compatibility. - Updated redhat-embeddedbitmap patch- Added xorg-ppc64-support-updates.patch, a forward port of ppc64 support from our XFree86 4.3.0 packaging - Made with_savetemps work on x86/alpha/ppc/ppc64/amd64, and remove -pipe when using -save-temps - Builds were failing with an mkfontscale SEGV or SIGILL on ppc/ppc64/s390 which has been tracked down to an aliasing bug in freetype 2.1.7. I have rebuilt freetype 2.1.7-3 with -fno-strict-aliasing which seems to work around the problem in testing. We will see if it fixes it once freetype is built. - Updated freetype dep in libs subpackage to 2.1.4 - Temporarily updated freetype-devel dep to 2.1.7-3 to force beehive to install it, so I get the -fno-strict-aliasing build - ifnarch ppc64/s390/s390x libI810XvMC.*- Initial xorg-x11 package, forked from the XFree86-4.3.0-63 package - Package name chosen as xorg-x11 due to lack of official upstream package name at this point. Name may change in the future, so people should not hard code any dependancies on it. Version and Release fields future proofed for similar reasons, as we do not know what upstream version will be. - Renamed StudlyCaps rpm macros XFree86CVSDate and CVSBuild to cvs_date and cvs_build to be more consistent in macro naming throughout specfile - Disabled the with_new_savage_driver, and with_via_driver options by default, as the included drivers are newer, although the via driver may lose DRI support temporarily which I may need to add back later - Disabled with_new_xft, with_new_xrender by default to use the stuff included in the tree, or use external Xft/Xrender packaging instead. - Force with_xterm to 0 for all builds as we have shipped xterm separately for a few OS releases now - Updated spec file License field to be more accurate "MIT/X11/others" - Updated freetype deps to "BuildRequires: freetype-devel >= 2.1.4", as 2.1.4 is what is currently included in the xorg sources and is the minimum needed to build properly - Replaced all hard coded Requires: XFree86-* with usage of %{name} macro instead, so package renames are easier to deal with - Dropped all patches that were previously backported from XFree86 CVS, or were already checked into XFree86 CVS since 4.3.0. - Temporarily disabled many patches that fail to apply, or fail to apply cleanly, for future investigation and cleanup, just to get rpms building - Ported the elfloader-linux-non-exec-stack patch to current xorg-x11 - Removed the Glide3 conditionalized BuildRequires hack that was previously implemented to work around Red Hat buildsystem problems - Disable removal of XIE/PEX docs as they are no longer provided in sources - Force the removal of _x11dir/src in %install, as the new tree installs the DRM source code there by default, and we do not want it because our DRM is in our kernel rpm. No sense confusing users with multiple DRMs. - Added getconfig bits to main server package - Added fonttosfnt to font-utils subpackage - Moved bdftopcf and bdftruncate to font-utils subpackage - All libraries are now supplied by default in upstream builds in both static and dynamic (.a and .so) formats. Libraries which are now supplied in shared form that were not in the past include: libFS.so, libGLw.so, libI810XvMC.so, libXRes.so, libXfontcache.so, libXinerama.so, libXss.so, libXvMC.so, libXxf86dga, libXxf86misc.so, libXxf86rush.so, libXxf86vm.so, libfontenc.so, libxkbfile.so, libxkbui.so - Added %ghost flagged fonts.scale files to 100dpi, 75dpi, cyrillic, and misc font directories, so that if these files are present, they are owned by the appropriate subpackages- Conditionalized with_voodoo_driver to not build under build_taroon- Removed kdrive subpackage, and all with_kdrive conditionalization, as it is just extra spec file noise currently, and we have never shipped it. Anyone building kdrive should instead be using kdrive from the xserver project on freedesktop.org - Removed all 'athlon' subarchitecture references from spec file as the rpm {ix86} macro includes it automatically. - Added XFree86-4.3.0-remove-copyright-symbol-to-fix-UTF-8.patch quick 2 second fix to correct manpage display under UTF-8 (#101243)- Built 4.3.0-62.EL with target build_taroon for Red Hat Enterprise Linux 3 erratum- Added new 2D driver from Alan Cox for Voodoo 1/2 hardware that does not require Glide in order to function, and provides RENDER acceleration and various other improvements above and beyond the XFree86 supplied "glide" driver which we have not shipped in the past. Driver version 1.0beta3. - Use new macros via_driver_name and voodoo_driver_name to avoid complexities in host.def with XF86ExtraCardDrivers define.- Added XFree86-4.3.0-keyboard-disable-ioport-access-v3.patch as a final patch without debug logging enabled, to fix (#115769) - Added spec file macro 'with_savetemps' for debugging purposes, disabling it by default. When used, it is set up to only get used on x86 for build_rawhide and build_psyche builds - Added XFree86-4.3.0-ati-ia64-no-nonpci-ioport-access.patch to fix ati driver issue on ia64 which causes IBM x455 system to machine check. Also added "#define ATIAvoidNonPCI YES" to host.def to activate this fix only on ia64 builds (#112175)- Added XFree86-4.3.0-keyboard-disable-ioport-access-v2.patch to try to fix (#115769) - Changed mkxauth to call chown as foo:bar instead of foo.bar as the latter syntax has been deprecated - Added XFree86-4.3.0-minor-typo.patch to fix a trivial typo that I spotted in an error message in linux int10 code. - Remove Buildrequires kudzu-devel, pciutils-devel, both of which were added on Mar 21, 2001 when Glide3 was included in the XFree86 packaging, but are no longer necessary. I detected this when the buildsystem failed my build due to being unable to meet the dependancy on kudzu-devel, and further investigation showed that dependancy is no longer necessary. - Added XFree86-4.3.0-xcursorgen-check-malloc-return.patch to make xcursorgen check the return codes on malloc before referencing allocated memory - Added XFree86-4.3.0-redhat-xcursorgen-do-not-build-included-cursors.patch to stop building the XFree86 supplied Xcursor cursors as we do not ship them- Added XFree86-4.3.0-xrandr-manpage-typo-fix.patch to fix manpage (#83702) - Added XFree86-4.3.0-radeon-9200-dvi-snow.patch to fix issue on Radeon 9200 when using DVI panel and encountering snow and other artifacts (#112073) - Updated XFree86-4.3.0-debug-logging-ioport-based-rate-setting.patch to have it patch both lnx_kbd.c and lnx_io.c because both files insanely contain identical cut and pasted copies of the exact same source code, so nothing shows up in the X server log when testing with previous patch as the calls never got invoked in lnx_kbd.c (#115769)- Added XFree86-4.3.0-debug-logging-ioport-based-rate-setting.patch to get some useful debugging information as to why the X server seems to be unable to use the KDKBDREP ioctl(), and is falling back to direct I/O port bashing, which causes problems in the kernel (#115769) - Removed /usr/X11R6/lib/X11/fonts/misc and /usr/X11R6/lib/X11/fonts/cyrillic from the default xfs font path, as /usr/X11R6/lib/X11/fonts/misc:unscaled is present by default, which is what we want, and the cyrillic one gets added automatically if cyrillic fonts get installed. - Updated libXrender to the version 0.8.4 maintenance release, which fixes a problem in 0.8.3 that caused antialiased fonts to be inadvertently disabled in Qt/KDE, and also caused color mouse pointers to appear as outlines on Radeon 7500 using Xinerama (#109351,109897) - Updated libXft to version 2.1.3 from freedesktop.org xlibs, and enabled it for build_rawhide and build_psyche just for testing purposes for now. - Conditionalize the patches XFree86-4.3.0-fixes-for-freetype-2.1.7-v2.patch, and XFree86-4.3.0-redhat-xft-loadtarget.patch to only be applied when with_new_xft is 0, as it is not needed for Xft 2.1.3 - Added XFree86-4.3.0-SDK-add-missing-includes-for-synaptics.patch from Paul Nasrat, to fix the SDK so that 3rd party drivers such as "synaptics" can be built in external rpm packages (#99351,103498,114804) - Comment out host.def supported driver list for alpha architecture, and just build the default drivers, as we no longer have official Alpha products, so there is no reason to restrict driver support to hardware that works or is reasonably considered to work. Ship it all and see what breaks. - Disabled tdfx driver DRI support in Fedora Core development, as we have at least temporarily deprecated Glide3 due to it no longer compiling properly. Alan was working on fixing this, so we will probably re-enable tdfx DRI support once Glide3 is working again. - Removed XFree86-4.3.0-fixes-for-freetype-2.1.7-v2.patch and replaced it with XFree86-4.3.0-general-fixes-for-freetype-2.1.7.patch which excludes Xft2 fixups, and XFree86-4.3.0-xft-2.1.2-fixes-for-freetype-2.1.7.patch which includes only Xft2 fixups needed for xft-2.1.2, as 2.1.3 includes this fix already- Oops, now we change DevelDriDrivers to DevelDRIDrivers so that the via DRI driver actually builds this time. - Oops, now we change '%define DevelDRIDrivers' to '#define DevelDRIDrivers' so that the via DRI driver actually builds this time. - Added XFree86-4.3.0-build-libXinerama-before-libGL-for-via-driver.patch because the via DRI driver links against libXinerama, and seems to be the only DRI driver which does so. This patch causes libXinerama to be compiled before libGL to ensure the via driver will link properly.- Happy VIAlentines Day Edition. Feel the love. - Updated the 'via' driver to latest driver from Alan Cox, which contains DRI support among other things. Only enabled for build_rawhide and build_psyche for the time being, until it has been adequately tested. - Added DevelDriDrivers to host.def with 'via' driver added, conditionalized to the with_via_dri_driver macro - Conditionalize with_via_driver to only build on x86 architecture - Added XFree86-4.3.0-i740-missing-symbol-vbeFree.patch to add a missing symbol to the i740 driver (#115676) - Built 4.3.0-56 with target build_rawhide for Fedora Core development- Added {x11datadir}/X11/xinit back to package list, which seems to have been inadvertently dropped during attempts to get package to compile on Red Hat Linux 9 s390 builds earlier this week. - Built 4.3.0-55 with target build_yarrow for Fedora Core 1 erratum - Built 4.3.0-55.EL with target build_taroon for Red Hat Enterprise Linux 3 erratum - Built 4.3.0-2.90.55 with target build_shrike for Red Hat Linux 9 erratum- Added XFree86-4.3.0-libXfont-security-CAN-2004-0083-CAN-2004-0084-CAN-2004-0106.patch to fix all recent security flaws in libXfont which are outlined in CAN-2004-0083, CAN-2004-0084, CAN-2004-0106, discovered by iDefense, David Dawes and others. This patch replace all previous libXfont patches from XFree86 builds 4.3.0-49 through to present. - Added XFree86-4.3.0-libXfont-security-CAN-2004-0083-CAN-2004-0084-CAN-2004-0106-v2.patch which is the same as the above patch, but modified to cleanly apply to 4.3.0, renamed to keep all patches present in src.rpm for comparative purposes. - Built 4.3.0-54 with target build_yarrow for Fedora Core 1 erratum - Built 4.3.0-54.EL with target build_taroon for Red Hat Enterprise Linux 3 erratum - Built 4.3.0-2.90.54 with target build_shrike for Red Hat Linux 9 erratum- Added XFree86-4.3.0-security-dirname-CAN-2004-0106.patch which replaces XFree86-4.3.0-security-fonts-alias-dirname3.patch, the new patch being the same but without the second hunk, as the patch Keith wrote for CAN-2004-0083 and CAN-2004-0084 already handled that issue so there was a conflict. - Built 4.3.0-53 with target build_yarrow for Fedora Core 1 erratum - Built 4.3.0-53.EL with target build_taroon for Red Hat Enterprise Linux 3 erratum - Built 4.3.0-2.90.53 with target build_shrike for Red Hat Linux 9 erratum- Added XFree86-4.3.0-security-fonts-alias-dirname3.patch in order to fix 2 additional buffer overflows in libXfont, discovered by iDefense and David Dawes. No CVE assignment has been provided yet. - Built 4.3.0-52 with target build_yarrow for Fedora Core 1 erratum - Built 4.3.0-52.EL with target build_taroon for Red Hat Enterprise Linux 3 erratum - Built 4.3.0-2.90.52 with target build_shrike for Red Hat Linux 9 erratum- Added XFree86-4.3.0-security-dirname-CAN-2004-0083-CAN-2004-0084-keithp.patch alternative patch written by Keith Packard, to fix CAN-2004-0083 and CAN-2004-0084 security issues - Added XFree86-4.3.0-security-fonts-alias-dirname-CAN-2004-0084.patch to the package, but disabled for now while we test the above patch from Keith Packard which addresses both security issues. - Built 4.3.0-51 with target build_yarrow for Fedora Core 1 erratum - Built 4.3.0-51.EL with target build_taroon for Red Hat Enterprise Linux 3 erratum - Built 4.3.0-2.90.51 with target build_shrike for Red Hat Linux 9 erratum- Fix issues detected in QA testing - Built 4.3.0-50 with target build_yarrow for Fedora Core 1 erratum - Built 4.3.0-50.EL with target build_taroon for Red Hat Enterprise Linux 3 erratum - Built 4.3.0-2.90.50 with target build_shrike for Red Hat Linux 9 erratum- Built 4.3.0-2.90.49 with target build_shrike for Red Hat Linux 9 erratum - Split {_x11datadir}/X11/etc/* glob previously wrapped using with_Xserver into a with_xterm portion and with_Xterm portion with the dir being always included, in order to work around obscure build failure on s390 on RHL 9. Yes this is an insane problem to have to fix because we do not ship an RHL 9 s390 product and never will. But we seek perfection however, and who knows, maybe next week we will release a Red Hat Linux 9 port to s390 for consumer desktops or something. - Rename with_included_xterm macro to with_xterm for naming consistency with other options, as it threw me off.- Built 4.3.0-49.EL with target build_taroon for Red Hat Enterprise Linux 3 erratum- Added XFree86-4.3.0-security-fonts-alias-dirname-CAN-2004-0083.patch to fix security issue in core fonts backend reported by iDefense in CAN-2004-0083 - Added build_maintainer_mode distribution version autodetection to simplify local build testing procedures, added dist_ver macro, dist_test parameterized macro (to keep jbj on his toes), and updated build_xxxx target autoconfig when build_auto_mode is enabled. This only affects local builds, not any Red Hat builds. - Enabled radeon-agp-detection-using-capability-list-walk patch on all builds, which was inadvertently left off on some due to misplaced macro conditional - Built 4.3.0-49 with target build_yarrow for Fedora Core 1 erratum- Rebuilt with build_taroon for RHEL 3 testing- Added XFree86-4.3.0-Xserver-dix-xkb-key-repeating-bug-CVS-backport.patch to fix a bug in DIX when xkb is being used that causes keys to repeat spuriously on some hardware under certain system loads. This patch has been backported from the 4.3.0-48 developmental head package. (#76959,114635) - Added XFree86-4.3.0-XRes-IncludeSharedObjectInNormalLib.patch to make libXRes get built PIC for bug (#114292) - Updated XFree86-4.3.0-missing-lib-sharedreqs.patch to remove dependancy on libXt caused by improper dependancy listing in SharedXmuuReqs (#113336)- Build test release for RHEL3 U2 testing- Temporary fork of 4.3.0-45 to add some patches for test builds, until post 4.3.0-45 (4.3.0-46 through 4.3.0-50) local-work-in-progress stuff is in clean enough shape for tree inclusion - Added XFree86-4.3.0-fixes-for-freetype-2.1.7-v2.patch so that XFree86 will build properly against freetype 2.1.7 (#114343)- Implemented new AGP/PCI autodetection in the Radeon driver by examining PCI configuration space and walking the PCI extended capabilities list in order to determine if the device implements the AGP capability. This code should work on _any_ AGP/PCI hardware generically and should be factored out into generic X server code in future XFree86 releases so all drivers can benefit from it. XFree86-4.3.0-radeon-agp-detection-using-capability-list-walk.patch should fix all Radeon PCI/AGP autodetection bugs, including (#111191). Some AGP Radeon users may experience a performance boost with this new driver if their card was misdetected and treated as PCI before, as pcigart mode works on AGP hardware, but is slower than using AGP. - Fixed build_rawhide to work the same as build_yarrow everywhere since the two are functionally identical for the time being.- Rebuilt 4.3.0-44 as 4.3.0-44.EL for RHEL3 QU1 update- Added XFree86-4.3.0-libfontenc-IncludeSharedObjectInNormalLib.patch to fix KDE build problem on AMD64 which links to the static libfontenc library and fails because it wasn't compiled with -fPIC, reported in bug (#111058) - Enable the open source vmware_drv.o video driver that ships with XFree86 on all builds now, to supply this driver as-is to users as a convenience although it is still unsupported by Red Hat. Users encountering video or other X related problems with this driver, need to report their problems directly to XFree86.org, or to VMware Inc. - Rebuild in rawhide for FC2 development- Added XFree86-4.3.0-nv-riva-videomem-autodetection-debugging.patch to be able to debug Riva TNT memory autodetection problems in the future (#109459) - Added new build_rawhide flag to wrap experimental changes and test patches with for Rawhide builds - Rename rpm macro from tlssubdir to _tlsdir, and enforce it's usage everywhere in the spec file- Rebuild 4.3.0-43 for Red Hat Linux 9 erratum with build_shrike set- Updated to XFree86-4.3.0-xf-4_3-branch-2003-11-03.patch to pick up latest fixes in the XFree86 4.3.x stable branch including: - Fix for crash on ia64 because of wrong setjmp buffer alignment (John Dennis) - Close freetype fontfile filehandle in mkfontscale, this prevents problems from limitation of simultaniously open files - Fixed erronous freeing of DisplayModeRec in xf86DeleteMode() when deleting the modePool in xf86PruneDriverModes() the 'prev' member has a different meaning for modePool modes than for ScrnInfoPtr->modes modes where it creates a doubly linked list - Fix some i830+ VT switch/exit crashes - Fix DRM_CAS on ia64 as used by the DRI (Bugzilla #778, John Dennis). - Removed XFree86-4.3.0-Xlib-XIM-bugfix-from-XFree86-bugzilla.patch, XFree86-4.3.0-ia64-setjmp-alignment.patch - Updated XFree86-4.3.0-ia64-drm-locking.patch as part of it is in the stable branch patch now. - Updated some spec file comments, and other mostly cosmetic changes. - Fixed some mistakes in spec file changelog dates.- Enable new Radeon support patches for shrike builds also to support newer Radeon hardware, so future erratum picks up these enhancements. - Backport XFree86-4.3.0-RandR-refresh-rate-rounding-error-fix-from-CVSHEAD.patch from CVS HEAD in order to fix bug (#108008) - Added XFree86-4.3.0-vidmode-SEGV-fix-from-CVS-HEAD.patch, backported from CVS HEAD to fix a SEGV in the vidmode extension (#101276) - Renamed build_cambridge target to build_yarrow to indicate the change from project name to final product name. - Added XFree86-4.3.0-rendition-complete-driver-backport-CVS20031031.patch which is a backport of the rendition driver from CVS head, including a couple bug fixes and the rest of changes are cosmetic. (#108693) - Disabled XFree86-4.3.0-rendition-disable-cause-of-SEGV.patch which should now be obsolete from above rendition driver backport.- This release is the long awaited answer to the meaning of life, the universe and everything. - Added XFree86-4.3.0-redhat-exec-shield-GNU-stack.patch to make the complete XFree86 build including Mesa et al. exec-shield friendly (arjanv, mharris) - Updated to new XFree86-4.3.0-Mesa-SSE-fixes-from-MesaCVS-v2.patch which should fix compatibility problems between DRI and 2.6.x kernels which were caused by the previous version of this patch. Linus reported the fix for this with details of the problem, and explanation of the solution, which I extracted out of CVS (#107932,106566,107829)- Added XFree86-4.3.0-ucs2any-C-implementation.patch which implements ucs2any in C, and is slightly faster than the ucs2any.pl perl script- More xfs initscript cleanups - Added XFree86-4.3.0-ia64-drm-locking.patch from John Dennis, to fix DRI locking deadlock on ia64 architecture (#104338,103936)- Fixed brown paper bag bug - missing "; then" in if clause of xfs initscript- Added conditional option with_bail_on_undef_syms which currently sets SharedLibraryLoadFlags to "-shared -Wl,-z,defs" in order to cause linking to fail completely when undefined symbols are found. Mesa compilation blows up, so I've disabled this option for now until I have proper time to spend on an investigation sometime in the future.- Updated xfs initscript to improve startup performance by not triggering the creation of fonts.dir due to new fonts.cache* files being found (#97240). - Updated xfs initscript to handle Opentype (otf, otc) fonts by calling mkfontscale + mkfontdir on them. - Updated xfs initscript to ignore all known font metadata file types when determining wether or not to trigger mkfontdir on non-truetype/non-opentype font file types. - Updated xfs initscript to run fc-cache if fonts.dir files get updated in any way. This is disabled for fc-cache version 1.0.2 which shipped on Red Hat Linux 8.0, as it SEGV's for some reason on that release, and 4.3.0 isn't officially supported for that release. - Updated xfs initscript to use "grep -qs" instead of output redirection in all grep invocations, as it is a bit cleaner code, possibly even faster. - Call mkfontscale in Syriac-fonts rpm post and postun scripts to properly generate fonts.scale and fonts.dir for these Opentype fonts. - Added XFree86-4.3.0-missing-lib-sharedreqs.patch to make sure libDPS, libXmuu, libXfont, libXp, libXpm, libXrandr, and libXv get linked to their dependant libraries when building XFree86, so that undefined symbols do not prevent prelinking from working. This solves other problems too, and now obsoletes the previous XFree86-4.3.0-missing-SharedXfooReqs.patch, and XFree86-4.3.0-libXrandr-missing-sharedreqs.patch patches as well (#106661)- Added XFree86-4.3.0-radeon-autodetect-pci-or-agp-cards.patch which is a backport from CVS of new code to autodetect wether a given card is AGP or PCI by reading PCI config space. This simplifies future updates to the radeon driver for future Radeon hardware. Older driver update patches can further be simplified later on to remove some unnecessary cruft from them. - Updated XFree86-4.3.0-radeon-support-backport-from-CVS-v2.patch to fix a few missing chip IDs in a section of radeon_driver.c - Enabled DRI support for ppc for build_cambridge target- This build contains conditionalized build changes for RHL 8.0, intended to allow pain free compilation and installation of 4.3.0 in RHL 8.0. All changes are conditionalized to build_psyche to limit their effects strictly to RHL 8.0 builds unless the change is an obvious no-risk fix that wont harm current development or erratum. - Renamed fontconfig-2.1-slighthint.patch due to a problem which was reported by a user recompiling XFree86 on Red Hat Linux 8.0, getting a fontconfig dependancy, recompiling and installing fontconfig, and then trying to recompile XFree86 from their previously installed XFree86 source rpm. Both XFree86 and fontconfig contain a patch of the same name, and the fontconfig one overwrote the one from the XFree86 packaging. The two patches are identical but the directory paths modified to match XFree86's fontconfig path. The new patch name is XFree86-4.3.0-redhat-fontconfig-2.1-slighthint.patch - Added "Provides: Xft-devel = %{xftver}" to XFree86-devel and "Provides: Xft = %{xftver}" to XFree86-libs, both wrapped with build_psyche for Red Hat Linux 8.0 compatibilty as many packages brokenly hardcode Xft dependancies - Disabled the Conflicts: Xft-devel and Conflicts: Xft lines from the -libs and -devel packages as they seem to prevent upgrades from working from XFree86 4.2.x to 4.3.0 on RHL 8.0 systems as the package both provides, and obsoletes, Xft and Xft-devel, and the 'conflicts' line seems to make it conflict with itself. - Fixed some broken with_fontconfig conditionalization for cases where XFree86 supplied fontconfig is used for building, just so the option actually works if someone sets it.- Added XFree86-4.3.0-ati-generic-shared-chip-data.patch to unify changes to atichip.h into a single harmless patch to avoid patch overlap and merge conflicts - Updated XFree86-4.3.0-radeon-support-from-ATI-backport-from-CVS.patch to remove changes to atichip.h as they're merged into the above patch instead - Added XFree86-4.3.0-radeon-support-backport-from-CVS.patch backport of support for newer Radeon 9600/9800/IGP/Mobility hardware from CVS head, along with a few minor bug fixes for Mobility and IGP. Very low risk change which is heavily audited, however currently configured to only build for cambridge and psyche until runtime tested sufficiently - Renamed XFree86-4.3.0-ati-radeon-dpms-on-dvi-v2.patch for consistency, to XFree86-4.3.0-radeon-dpms-on-dvi-v2.patch - Added XFree86-4.3.0-Xserver-xf86PciInfo-updates.patch which from now on will hold all xf86PciInfo updates. Moved all Radeon, savage, and S3 updates from other patches into this file.- Rebuilt 4.3.0-35 as 4.3.0-35.EL for RHEL- Updated to XFree86-4.3.0-xf-4_3-branch-2003-10-02.patch to pick up latest fixes in the XFree86 4.3.x stable branch including: - xdm build fixes from previous security update - Use pam_strerror() to print an error message after pam_setcred() fails, C style unification - xdm portability fix - Added XFree86-4.3.0-Xlib-XIM-bugfix-from-XFree86-bugzilla.patch to fix an XIM bug introduced into XFree86 CVS in late August, reported in XFree86 bugzilla, fixed in the trunk but not the 4.3 branch yet. (#106058) - The XIM bug above was introduced in XFree86-4.3.0-xf-4_3-branch-2003-09-11.patch which the spec file claims I added Sat Sep 6, 2003 in version 4.3.0-30, which would mean that release was affected also, however this isn't the case. A deeper investigation shows that the changelog got the patch added, but the actual patch got misplaced somehow, so didn't get applied until 4.3.0-32. As such, I'm adding an updated comment to the 4.3.0-32 changelog entry to reflect this goofup.- Made 'hostname' use -f in BuilderString, so that the FQDN of the buildhost is present in X server startup log info - Updated server startup vendorstring for Fedora Core, RHEL, and to indicate Red Hat Linux distro releases 8.0 and 9- Added XFree86-4.3.0-savage-scaling-bz274.patch to fix (XF86 #274) - Added XFree86-4.3.0-savage-memleak-plug-bz278.patch to plug memory leak in the savage driver (XF86 #278) - Added XFree86-4.3.0-savage-memleak-plug-bz279.patch to plug memory leak in the savage driver (XF86 #279)- Updated to XFree86-4.3.0-xf-4_3-branch-2003-09-26.patch to pick up new security fixes from CVS - Updated XFree86-4.3.0-redhat-libGL-exec-shield-fixes.patch to new patch XFree86-4.3.0-redhat-libGL-exec-shield-fixes-v2.patch which reorders some includes in mem.c so it builds. Still cambridge only. (#104029)- Added XFree86-4.3.0-redhat-libGL-exec-shield-fixes.patch experimental patch to cambridge only from John Dennis to fix libGL / DRI / exec-shield problems with executable page mappings (#104029,91784,101647) - Added XFree86-4.3.0-redhat-libGL-TLS-buildfix.patch to fix the libGL opt patch Imake TLS stuff. Cosmetic only.- Created XFree86-4.3.0-Xserver-dualhead-segv-with-no-busid.patch to fix a problem that Roland McGrath was experiencing, where the X server will crash when starting on a system with one dualhead video card, with the config file omitting the BusID keywords. The crash is due to a format string bug in xf86Helper.c during startup time in a codepath that is rarely executed. - Moved new Xft/Xrender tarball unpacking and preparation from build section to prep section where it belongs, in order for patches we apply to xft et al to not get missed, and also so rpmbuild -bp results in something closer to the actual sources we are using. There are a few more areas needing tidying in this regard also.- Rebuilt 4.3.0-32 with build_taroon set as 4.3.0-32.EL in order to pick up the libGL patch in Taroon- Updated to XFree86-4.3.0-xf-4_3-branch-2003-09-17.patch to pick up latest xdm security fixes from CVS stable branch "xf-4_3-branch": - Backport xdm code from -current, including following changes: - use better pseudo-random generators to generate magic cookies, including EGD-like prng daemons support. - add support for LISTEN keyword in Xaccess - deal with small memory leaks - use SIOCGLIFCONF to query network interfaces where available - fix XDMCP bug that could cause localhost entries in /etc/X0.hosts to be lost - add xdm /dev/random handling for Solaris - fix XDMCP queries on systems using getifaddrs(). - Fix for http://www.cve.mitre.org/cgi-bin/cvename.cgi?name=CAN-2003-0730 - Enabled the libGL optimization patch for Taroon, but disable TLS support as it is experimental and only intended for Cambridge builds. - Updated XFree86-4.3.0-redhat-radeon-forcelegacycrt-for-alan.patch to remove CVS revision strings from the patch so it applies cleanly with new update - Dropped XFree86-4.3.0-xaa-pixmap-cache-crash-clash.patch as the stable branch patch update now has this fix- Added "export LANG=C" to the build and install sections of the spec file in hopes it will speed up all script processing during building. - Added xrender.pc pkg-config support using an expansion disabled here-document with rpm macros- Rebuilt for Taroon- [CHANGELOG UPDATE] - the XFree86-4.3.0-xf-4_3-branch-2003-09-11.patch referenced below originally, actually got left out of this build inadvertently, and the changes referenced did not get included until the 4.3.0-32 build above. - Updated to XFree86-4.3.0-xf-4_3-branch-2003-09-11.patch to pick up latest fixes from CVS stable branch "xf-4_3-branch", including: - Fix calculation of CRTC2 frame offset with page flipping in radeon driver (Michel Daenzer) - Fix an XAA pixmap cache server crash that can happen in some cases when the off-screen memory is heavily fragmented (David Dawes, based on #5752, Koike Kazuhiko, Chisato Yamauchi). - Fixed a crash when _XIMProtoOpenIM(), hich is called through XOpenIM() API when protocol IM is being set up, fails (Bugzilla #618, Hisashi MIYASHITA). - Don't call FBIOPAN_DISPLAY ioctl with arguments that will cause a confusing if harmless error (Michel Daenzer) - Fixes for potential integer overflows in font libraries. (blexim, Matthieu Herrb). - For integer overflow tests, use SIZE_MAX which is more standard than SIZE_T_MAX, and default to ULONG_MAX for the case of LP64 systems. Based on reports by Matthias Scheler and Alan Coopersmith (Bugzilla #646). - Created a new XFree86-modules-unsupported package, and moved some of the modules which are unsupported and provided totally as-is into this new subpackage - Disabled SDK-add-missing-includes-for-vnc patch as it causes the X server build to fail, reverted to previous patch for synaptics. - Added XFree86-4.3.0-ia64-setjmp-alignment.patch to fix an issue with setjmp on ia64 which results in an X server crash when using Sun Java. Fixed by John Dennis for bug IDs (#103610,104191) - Remove {_x11datadir}/X11/xserver on builds disabling with_Xserver, as a dead symlink was found by with with_dead_symlink_test. - Modified with_dead_symlink_test to be cleaner.- Enable libOSMesa on s390/s390x/ppc/ppc64, as it was disabled but some of the files were shipped anyway, and there's no really good reason to not supply it that I can think of. - Add a new conditionalized test (with_dead_symlink_test) to test for dead symlinks to the end of the install section of the specfile and enable it by default. - Added Requires: xterm to the XFree86-twm subpackage because twm configuration requires xterm by default. This also will fix another problem where upgrading the XFree86 packages from from RHL 9 or earlier to XFree86 after xterm was split out into it's own package, will cause xterm to be lost unless the user manually installs the xterm package. Now, as long as they have XFree86-twm installed, xterm will be dragged in as a dependancy, since it is one. If anyone knows of similar xterm dependancies either in other XFree86 subpackages or in other Red Hat Linux packages, they should file a bug report and add me to the CC list. - Added XFree86-4.3.0-SDK-add-missing-includes-for-vnc-and-synaptics.patch as an updated of the SDK patch for the GPL synaptics driver (#99351)- Move perlification of xft.pc earlier so it has effect - Fix broken symlinks for libXrender.so.2 similar to the prior broken Xft libs- fix typo in perl script- Build 4.3.0-27, release name (Conquest) - Added XFree86-4.3.0-SDK-add-missing-includes-for-vnc.patch, contributed by Paul Nasrat in order to help get VNC compiling with the SDK (#103824) - Added XFree86-4.3.0-radeon-disable-VideoRAM-option.patch which causes the Radeon driver to ignore any user supplied VideoRAM option in the config file, as the radeon driver properly autodetects RAM on all supported video hardware, and the option to override it is misused by users who think they have more RAM than they do, resulting in the creation of a problem where none existed. In the unlikely case a model of radeon is found that does not autodetect the amount of videoram properly, we can easily fix that with information provided by ATI at that point in time. - Fixed broken libXft symlinks, hopefully got it right this time. - Fix xft.pc file by munging @VERSION@ into the proper version- Build 4.3.0-26, release name (Brazil Nut) - Added XFree86-4.3.0-xaa-pixmap-cache-crash-clash.patch to fix bug in XAA pixmap cache code which could cause a SEGV. Patch by Thomas Winishhofer, submitted to me by Daniel Stone at Debian. - Added XFree86-4.3.0-nls-cjk-utf8.patch fix for (#101377) - Added XFree86-4.3.0-Mesa-SSE-fixes-from-MesaCVS.patch fix for SSE/3Dnow detection problem, backported from Mesa3D CVS tree, not present in XFree86 CVS. - Added XFree86-4.3.0-nls-compose-en_US.UTF8-lxo.patch to fix problem with us_intl keymap and funny i18n characters (#80244) - The XFree86-4.3.0-redhat-xft-loadtarget.patch was getting applied to the Xft lib that ships with XFree86 during the patch phase, but we immediately replace Xft with the latest Xft2 in the build section thus losing the patch, so I've added code in the build section to re-patch the new Xft2 - Put the xft.pc pkg-config file back into the filelist - Fix libXft.so.2 symlink to libXft.so.2.1.1 instead of libXft.so.2.1 (#103636)- The Australian special edition release, codename (Jumbo Tiger Shrimp) - Re-enabled with_new_xft_and_xrender, this time using Imake, dealing with limiations of Imake by renaming the generated libraries to 3 digit versions- Disable Glide3 support on ia64 as it does not really work anyway - Disable xkb-multilayout-modifier-fix-CVS11137 as the ALT-TAB problem is back now, and this is the next likely candidate cause- Updated XFree86-4.3.0-redhat-libGL-opt.patch from Jakub , so that libGL TLS support now works on all x86 class CPUs, rather than just i686 and higher. New version is: XFree86-4.3.0-redhat-libGL-opt-v2.patch - Disabled xkb-us_intl-and-other-fixes-CVS11216 and xkb-fix-sticky-modifiers patches since they fix one bug, to create another, and fix that bug to create several more. The original problem was much less worse and affected far less people. - Enabled with_debuglibs temporarily for debugging purposes- Added --libdir=%{_x11libdir} for Xft/Xrender stuff- Change ./configure lines of Xft/Xrender to have --prefix=/usr/X11R6- This package only compiles on Linux, and is not intended for use on other OSs. Made explicit now. - Test build with SDK enabled to try to troubleshoot- Re-enabled new Xft/Xrender in order to try and debug why the builds are failing ONLY in the Red Hat buildsystem.- Added XFree86-4.3.0-libGL-multitexture-defines.patch to fix bug where wine and various other OpenGL applications wont compile with Mesa from 4.3.0 due to a gl.h header file bug introduced in 4.3.0 (#97840) - Added XFree86-4.3.0-libXrandr-missing-sharedreqs.patch to fix bug (#102924) - Added XFree86-4.3.0-xkb-fix-sticky-modifiers.patch to fix problem in XFree86 rpm releases 4.3.0-19 through 4.3.0-21 which cause ALT-TAB and some other modifier combinations to break. Bug (#102668) - Moved all xkb datafile patches to section 800-829 in specfile - Added conditional BuildRequires: glibc >= 2.3.2-66 and Requires: lines if with_GL_TLS libGL optimization patch is enabled, to ensure the system has the required glibc installed. (#102175) - Disabled the new Xft and Xrender code by toggling with_new_xft_and_xrender, because of an unexplainable build failure that has started occuring only in the Red Hat buildsystem, which I can not reproduce locally. 4.3.0-21 built perfectly fine in beehive, but the current rpm fails when compiling the new Xft, saying that ld can't find -lXrender. I've absolutely no clue why this is occuring, all of the builds compile both on x86 and AMD64 locally here, and am unable to debug it on the systems it occurs on in the buildsystem.- Added XFree86-4.3.0-libXpm-missing-header.patch from John Dennis to fix bug (#101017) - Added XFree86-4.3.0-ia64-slowbcopy2.patch by John Dennis to replace the earlier patch XFree86-4.3.0-ia64-slowbcopy.patch which fixes a semi obscure problem in slowbcopy (#98139) - Integrated Keith Packard's upstream Xft-2.1.2 and Xrender-0.8.3 bugfix packages from fontconfig.org into XFree86 4.3.0. This fixes a variety of Xft/Xrender bugs scattered throughout bugzilla against many GNOME and other applications, etc. (#99468, 99469)- Disabled patches XFree86-4.3.0-revert-en_US.UTF-8-Compose.patch and XFree86-4.3.0-en_US.UTF-8-Compose-safer.patch now to provide stock XFree86 en_US.UTF-8 compose table. It had been disabled days before Red Hat Linux 9 went gold in order to minimize the risk factor of the change due to lack of widespread beta testing. Upstream X has not received bug reports about any problems caused by the change so it seems safe. It also should provide enhancements to some subset of users, and might fix some problems that have been reported to us, which might not be present in stock 4.3.0. Now is a safer time to enable this change than 3 days before our gold release of RHL 9. ;o)- Added XFree86-4.3.0-xkb-broken-modifiers-be-am-CVS10956.patch to fix modifier bug in be and am keymap files (#87679) - Added XFree86-4.3.0-xkb-it-keymap-update-CVS10957.patch to fix bugs in Italian keymap files (#88080) - Added XFree86-4.3.0-xkb-hr-keymap-fixes-CVS10967.patch to fix a small issue in the hr keymap files, and add a new hr_US keymap. - Added XFree86-4.3.0-xkb-broken-capslock-fix-CVS11006.patch to fix bug in DVORAK (and other) keymap files which breaks capslock. (#87494) - Added XFree86-4.3.0-xkb-ru-group-number-fix-CVS11092.patch to fix broken group number in russian map - Added XFree86-4.3.0-xkb-ua-add-missing-letters-fix-CVS11134.patch to add missing letters to Ukranian keyboard map - Added XFree86-4.3.0-xkb-multilayout-modifier-fix-CVS11137.patch workaround for problems that arise when in multi-layout map different modifier keysyms share the same key - Added XFree86-4.3.0-xkb-it-small-fix-CVS11150.patch which contains a small fix for the Italian layout - Added XFree86-4.3.0-xkb-us_intl-and-other-fixes-CVS11216.patch to add Euro sign support to us_intl keyboard map. Patch also fixes various other xkb bugs. (#91630)- Disabled SDK temporarily as it does not work on all architectures - Added XFree86-4.3.0-mouse-draglock-fix-CVS10885.patch to fix mouse draglock issue with buttons greater than 4 - Added XFree86-4.3.0-restore-underline-position-in-freetype-CVS10891.patch which should fix some issues with mozilla in bugzilla - Added XFree86-4.3.0-fbdev-mode-validation-segv-fix-CVS10893.patch to fix SEGV in fbdev's mode validation - Added XFree86-4.3.0-vidmode-memleak-fixes-CVS10895.patch to fix various memory leaks and uninitialized struct fields in vidmode extension- Enabled the XFree86-sdk subpackage conditional so we have the SDK- Added Jakub Jelinek's libGL optimization patch, which allows parts of libGL to be built PIC without affecting performance, thus allowing apps linked to libGL to be prelinked (XFree86-4.3.0-redhat-libGL-opt.patch) as well as other optimizations - Toggled with_debuglibs off so X doesn't take 10 years to compile- Rebuilt 4.3.0-16 as 4.3.0-16.EL for Red Hat Enterprise Linux development- Updated to XFree86-4.3.0-xf-4_3-branch-2003-07-09.patch to pick up latest fixes from CVS stable branch "xf-4_3-branch", including: - Fix repeated image problem when using a vesa video mode before starting the Xserver on the S3 Trio3D cards - Fix for a crash if a scalable font has a bitmap entry in fonts.dir (#5687, Bugzilla #332, Juliusz Chroboczek). - Fix for xfs crashes in Freetype backend (Bugzilla #242, Juliusz Chroboczek). - Pull twm fixes (signal handler, empty windows menu) from CVS head - Added XFree86-4.3.0-libXi-freeze-fix-CVS11329.patch to fix a deadlock in libXi where is called _XLockDisplay() twice when calling an Xi function that calls XGetExtensionVersion() (Bugzilla #260, Bastien Nocera, Owen Taylor) - Added XFree86-4.3.0-ps2-mouse-resolution-fix-CVS11329.patch to change PS/2 mouse resolution to 8 counts/mm as some broken mice have problems with other values (Windows uses 8) - Added XFree86-4.3.0-xkbuse-avoid-roundtrip-CVS11329.patch to correct a precedence problem in test in xc/lib/X11/XKBUse.c XkbUseExtension(). This avoids an extra round trip during application startup (Bugzilla #473, Owen Taylor) - Added XFree86-4.3.0-ia64-slowbcopy.patch by John Dennis to insert a short delay in the xf86SlowBcopy() routine and prevent PCI bus lockups on ia64 which were observed with the nv driver (#98139)- Added XFree86-4.3.0-via-driver-cvs20030611.tar.bz2 snapshot from XFree86 CVS to add support for the VIA EPIA onboard Castlerock CLE266 video chipset- Removed "#!/bin/bash" from top of specfile as midnight commander 4.6.0 now has built in rpm specfile highlighting! YaY! - Added XFree86-4.3.0-mga-enable-video-rom-before-using.patch to fix long standing bug in mga driver which causes an MCA on Alpha and likely other platforms as well (#91711)- Rebuild 4.3.0-14 as 4.3.0-14.EL for Red Hat Enterprise Linux development- Added XFree86-4.3.0-ia64-pci-infinite-loop.patch fix for a problem that occurs while doing a pci scan on Intel Tigers which leaves pci data structure with multiple root buses (primary == -1) causing an infinite loop, for now exit the loop when primary < 0 - John Dennis - Remove X server symlink check from xfs initscript for bug (#54630)- Updated to XFree86-4.3.0-xf-4_3-branch to pick up the latest fixes from the stable XFree86 4.3.0 branch including - Fix programming error in ix86 motherboard chipset determination - Workaround for broken devices that do not implement the header type field in their PCI configuration space - Fix infinite loop that occurs on systems whose PCI configuration space does not advertise a host bridge - Fix to prevent PCI and CardBus resets when VT switching - Set Mesa hooks to flush vertices on state changes in Radeon 3D drivers, which I believe fixes (#91784) - Fix lockup on server reset in radeon driver - Check for NULL tObj in the i830 3D driver TexEnv function. This fixes a FlightGear crash - Prevent a SIGFPE with the glint/pm3 driver when attempting to display an XVideo image less than 8 pixels wide - Fix a SEGV that can happen with Riva128 cards - Fix SiliconMotion driver for mode switching and SEGV problem when initializing Xv functionality (XF86 #50) - i810/815 depth buffer needs to be a multiple of the tile size. This fixes 3D corruption near the bottom of the screen at 800x600@16bpp (XF86 #283) - Fixed Imakefiles so the driver SDK builds again - Fix XDMCP queries in xdm on systems using getifaddrs() (XF86 #277) - Fixed xfs rpm preinstall script to ingore errors and redirect all output to /dev/null. This was accidentally removed in a prior build, and causes xfs installation to fail. (#91822 & 91706,91707,91733,91756) - Added XFree86-4.3.0-savage-bugfix-from-xf-4_3-branch-2003-05-28.patch which is extracted from xf-4_3-branch and reapplied on top of our 1.1.27mh driver since we have updated this driver but do not want to lose fixes coming from the stable branch. - Dropped patches that are now included in the stable branch: XFree86-4.3.0-radeon-irq-gen-lockup-from-cvs.patch, XFree86-4.3.0-radeon-dualhead-cursor-crash-from-CVS.patch - Disabled XFree86-4.3.0-siliconmotion-Xv-stability-fix.patch as it fails to apply now. Pending future investigation as it is non-critical anyway. - Added XFree86-4.3.0-redhat-lock-version.patch, enabled it, and locked down the version to 4.3.0 (currently stable branch is 4.3.0.1)- Enabled with_debuglibs now so that the libraries are debuggable during development, since it is not possible to have .debuginfo packages for X for various complex technical reasons, and it wont likely be possible for quite some time unfortunately. The XFree86-libs package will get much larger due to this change, so I may have to revert it if people complain, or if it slows down anything. This is a *TEMPORARY* solution, and will be removed sometime. - HasFontconfig seems to work now, so I am using it instead of UseFontconfig to see if anything breaks in rawhide - Removed with_freetype2 macro, and all conditional code using it throughout the specfile, as we never want to include the XFree86 freetype2 - Added missing Provides: xpm back to XFree86-libs package which got sideswiped due to the conditionalized with_freetype2 junk during development - Made triggerin script conditionalize referencing rstart dir- Disabled rstart, as it is probably not even used nowadays. People should be using ssh/rsh or some other modern mechanism for running remote X apps. - Change UseFontconfig to HasFontconfig to test if it works- Define "XF86Server NO" on s390, s390x, ppc64, as we do not ship X servers on these platforms. We use a 32bit X server on ppc64. - Wrap all X server modules, docs, config files, man[45]/*, include/*.h with with_Xserver in file lists so that things are properly packaged on the archs that we do not ship Xserver on (s390/s390x/ppc64)- Update XFree86-4.3.0-ppc64-support-v3.patch to -D__powerpc__ on ppc64 or Radeon driver explodes while building - Remove -fno-strength-reduce from x86 and x86_64 host.def optimization flags - Added ppc64, ia64 host.def optimization flag overrides, defaulting them to their existing settings- Updated XFree86-4.3.0-elfloader-linux-non-exec-stack-v2.patch to fix elfloader to page align addresses it passes to mprotect- Disable xterm from building, as it is now shipped in its own separate package - Build 4.3.0-9 in rawhide and 4.3.0-9.0.EL for RHEL- Added XFree86-4.3.0-Xserver-includes-ansify-prototypes.patch to ansify remaining function prototypes in the X server include dir - Added XFree86-4.3.0-ppc64-support-v2.patch which contains some more complete updates from SuSE for PPC64, replacing XFree86-4.3.0-ppc64-support.patch - Set HasAgpGart NO for ppc64 - Added new macro with_Xserver to conditionalize the inclusion of the X server, driver and extension modules, and define it to 0 for s390, s390x, ppc64 - Updated file lists to use with_Xserver macro instead of ifnarch s390/s390x- Do not install or ship mmapr, mmapw, pcitweak, scanpci, libOSMesa.so.4.0, pcitweak.1x.gz, or scanpci.1x.gz on ppc64, s390 or s390x- Enable XFree86-4.3.0-elfloader-linux-non-exec-stack.patch again, and remove ifnarch x86_64 on XFree86-4.3.0-xlib-dual-malloc-memleak-lcPrTxt.c.patch and XFree86-4.3.0-craptastic-cast.patch - Disable XFree86-4.3.0-missing-SharedXfooReqs.patch temporarily as it breaks xauth + ssh on s390, and thus requires deeper investigation - Build 4.3.0-8 for rawhide, and 4.3.0-8.RHEL.0 for RHEL devel- Replaced XFree86-4.3.0-ia64-new-elfloader-relocations.patch with a new version that John added to fix (#89976) - Added XFree86-4.3.0-radeon-ia64-preint10.patch from John Dennis, to solve a problem which is yet to be announced. ;o) Will update comment once I have details.- Added XFree86-4.3.0-ia64-new-elfloader-relocations.patch for ia64 ELF relocations for http://bugs.xfree86.org/cgi-bin/bugzilla/show_bug.cgi?id=195 - Replaced XFree86-4.3.0-missing-SharedXfontReqs.patch with a new patch XFree86-4.3.0-missing-SharedXfooReqs.patch that solves similar problems for many other X libraries on Linux, and should fix (#89860)- Apply XFree86-4.3.0-xf-4_3-branch-2003-04-26.patch to update to the latest fixes from the xf-4_3-branch stable branch of XFree86 CVS - Removed XFree86-4.3.0-radeon-R300-fix-pll-value-from-cvs.patch as it is incorporated into the stable branch above: - Added XFree86-4.3.0-libXfont-null-pointer-dereference-fix.patch which fixes a null pointer dereference in libXfont - Added XFree86-4.3.0-siliconmotion-Xv-stability-fix.patch to fix siliconmotion driver for mode switching and SEGV problem when initializing Xv functionality - Added XFree86-4.3.0-Xserver-non-pc-keyboard-fix.patch to fix problem on non-PC keyboards caused by changes to make SysRq generate the same keycode as PrtScrn - Added XFree86-4.3.0-nv-graphic-engine-setup-reset-fix.patch to stabilize VT switching and server startup/shutdown - Added XFree86-4.3.0-nv-new-chip-ids.patch which adds new PCI IDs for GeForce FX and Quadro FX - Added XFree86-4.3.0-nv-riva-dualhead-fixes.patch to fix problems that can arise on newer chipsets on dualhead configurations - Added XFree86-4.3.0-nv-filter-imagerect-pixel-range.patch to fix a problem where the HW filter would include pixels outside of the image rect being displayed. - Added XFree86-4.3.0-nv-dpms-workaround-for-broken-monitors.patch - Added XFree86-4.3.0-nv-unresolved-symbols.patch to add missing vgaHWUnmapMem - Added XFree86-4.3.0-rendition-disable-cause-of-SEGV.patch - Added XFree86-4.3.0-trident-half-a-jittered-screen.patch to fix bug in the trident driver that caused old cyber 9382/9385 chipsets to display a half jittered screen - Added XFree86-4.3.0-Xserver-limit-440EX-440LX-to-32-pci-busses.patch - Added XFree86-4.3.0-xaa-trapezoid-filling-bug.patch in a rare 8x8 pattern fill case in XAA generic code - Added XFree86-4.3.0-Xserver-Xineramify-Xscreensaver-extension.patch to make screensavers work in dualhead Xinerama mode - Added XFree86-4.3.0-Xlib-XOpenDisplay-SEGV-fix.patch to fix many places in Xlib which could segfault due to not checking to see if it has received as much data as needed, causing it to potentially read uninitialized memory - Added XFree86-4.3.0-mga-manpage-overlay-option.patch which better documents the Matrox Overlay option as many people are confused by this - Added XFree86-4.3.0-r128-unresolved-symbols.patch to add various symbols missing from symbol lists that generate errors when DRI is disabled- Removed XFree86-4.0.1-alpha-pcibus-lemming.patch as it is not clear what it actually does, is not documented anywhere what if any bug it fixes, and nobody remembers what it was for, and we have not applied it for quite a while now either. - Added XFree86-4.3.0-radeon-irq-gen-lockup-from-cvs.patch which works around a lockup that can occur on server shutdown/restart with the radeon driver due to bugs in DRM IRQ handling in the kernel - Added XFree86-4.3.0-radeon-support-from-ATI-backport-from-CVS.patch which is an amalgamation of patches done by Hui Yu at ATI which implement: Radeon IGP320/330/340 support (2D only), RV280 (9200/M9+) support, Fix for some M9 laptop panels, Improved version of monitor detection code, Fixed bug for two or more radeon cards, Man page updates, Workaround for double scan modes problem at high resolutions, Overlay scaling problem when RMX is used, and also PPC updates for Radeon from Michel Daenzer - Replaced XFree86-4.2.99.902-ati-radeon-alan-legacy-vga.patch with new XFree86-4.3.0-redhat-radeon-forcelegacyvga-for-alan.patch reimplementation around the new radeon driver code. Needs testing by Alan.- Enable _unpackaged_files_terminate_build & _missing_doc_files_terminate_build options and clean up remaining file failures so we are swooshy clean now - Update %install file deletion list for s390/s390x/ppc - Moved DELETE FILES section after STRIP FILES section in specfile - Added XFree86-4.3.0-redhat-xlib-linux-fix-avoiding-substance-abuse-job.patch as a hideous kludge to work around the even more ugly hideous Solaris broken weirdness junk triggering off _LP64 in XlcDL.c- Renamed TinyX subpackage to kdrive - Made all kdrive X servers setuid root in filelists- Added XFree86-4.3.0-xbiff-file-heirarchy-standard.patch for FHS 2.2 compliance of xbiff - When not shipping config tools, defining "BuildXFree86ConfigTools NO" in host.def speeds up build - Added XFree86-4.3.0-disable-building-apps-we-do-not-ship.patch to disable building applications we do not ship such as xedit, xman, xmh - Moved include/pixmaps from XFree86-devel to XFree86-xf86cfg where it belongs, although this stuff really should not be in _x11includedir- Added PPC64 support with XFree86-4.3.0-ppc64-support.patch- Added XFree86-4.3.0-ati-radeon-dualhead-cursor-crash.patch to fix radeon dualhead crash problem when mouse cursor crosses screens (#87854,88748)- Added XFree86-4.3.0-missing-SharedXfontReqs.patch to try and fix (#88355)- Added XFree86-4.3.0-redhat-bug-report-address-update.patch which changes the X server bug report messages on where and how to file bug reports- Added XFree86-4.3.0-ati-r128-Xv-ecp-divisor.patch which fixes a problem on Rage 128 in which high resolution images displayed with the XVideo extension have weird "green" noise in them (#88033) - Added XFree86-4.3.0-makefile-fastbuild.patch to bypass the main Makefile clean and depend targets when make World is ran- Added XFree86-4.3.0-xlib-dual-malloc-memleak-lcPrTxt.c.patch which fixes a memleak in Xlib caused by double malloc. Not being applied to x86_64 yet, waiting for approval to fix (#86108) - Added XFree86-4.3.0-craptastic-cast.patch to non-x86_64, just a cosmetic casting fixup of (off_t)(off_t) to just (off_t)- Updated XFree86-4.3.0-glint-unresolved-symbols.patch to add several more missing DRM symbols that another user reported when DRI is not used. This compliments the existing symbol name additions with an even more complete list, and should fix the problems reported in bug (#85191) as well as adding to the bugfix for (#85805) - Added XFree86-4.3.0-elfloader-linux-non-exec-stack.patch to patch the XFree86 ELF loader to work properly on Linux with a kernel with a non-exec stack, however this patch is disabled on x86_64 and possibly other architectures for now- Added XFree86-4.3.0-ati-radeon-R300-fix-pll-value.patch which fixes a bug in the Radeon driver (on all archs) that occurs only on R300 chips (Radeon 9500/9700) due to a broken PLL define in radeon_regs.h (#85784) - Re-enabled XFree86-4.3.0-glint-unresolved-symbols.patch on x86_64 as it is approved now by msw.- Modified xdm.pamd and xserver.pamd to remove explicit paths with hard coded /lib dir, since it is in /lib64 on lib64 archs, and so pam console ownership is screwed without this change. Similar problems were fixed in September in many other packages, but no bug reports were filed in bugzilla so I was aware of this problem, so it remained broken until I just discovered it myself while trying to debug a screensaver related issue (#85795) - I have made the XFree86-4.3.0-glint-unresolved-symbols.patch patch added in 4.3.0-2.1 ifnarch x86_64 to exclude this patch from x86_64 builds, even though it is an obviously correct and harmless bugfix on all architectures and should be applied (#85805)- Added XFree86-4.3.0-glint-unresolved-symbols.patch to fix some unresolved drm symbols in the glint driver. Possible fix for (#85191)? - Removed unused patch - XFree86-4.0.99.3-Xwrapper.patch (#85336) - Renamed xdm-4.0-servonly.patch to XFree86-4.0-xdm-servonly.patch- Removed some junk comments from specfile, the commented out old cursors and xtrap-clients subpackage junk and a few other things - Corrected a mistake in OTF font directory file list that was named TTF instead of OTF, causing a ghost file to be missed - Verified all file lists to be accurate just to doublecheck the 4.3.0 release is not accidentally missing any files - Removed mharris_noclean_mode developmental cruft from specfile, as I do not use it in my developmental builds anymore and wont need it after transitioning to _unpackaged_files_terminate_build mode - Removed XFree86-4.2.99.901-012_r128_xv_btyeswap_fix.patch as current CVS fixes the problem in a different way, and this patch reverts that - Removed the changes deemed potentially unsafe from en_US.UTF-8 compose table to have minimalistic changes- Updated to official XFree86 4.3.0 release via CVS export of xf-4_3_0- Updated CVS snapshot to 4.2.99.902-20030226, manually checked each diff and verifying each code change is safe. No questionable changes today.- Updated CVS snapshot to 4.2.99.902-20030225, manually checked each diff, accept all documentation updates, verified each change. - Added XFree86-4.3.0-revert-en_US.UTF-8-Compose.patch to back out changes made in CVS to en_US.UTF-8 Compose tables as they can not be easily verified to be 100% correct and regressionless at this point - Removed XFree86-4.2.99.902-xft-version-bump.patch as it is in CVS update- Added XFree86-4.2.99.902-xft-version-bump.patch to fix Xft library .so version from 2.0 to 2.1 since it had an addition of UTF-16 APIs. This fixes http://bugzilla.mozilla.org/show_bug.cgi?id=175108- Updated my radeon XFree86-4.2.99.902-ati-radeon-dpms-on-dvi-v2.patch patch to play things a bit safer by removing MT_LCD. There is a more appropriate and robust way to do this that I will examine in the future. - Removed unused WithNewCyrixDriver crud from specfile - Removed unused with_bk crud from specfile - [SECURITY] Added XFree86-4.3.0-xterm-can-2003-0063.patch to fix xterm titlebar issue reported in CAN-2003-0063 by disabling the feature that allows the problem.- Updated CVS snapshot to 4.2.99.902-20030224, which makes another cursor glitch fix to generic code, and fixes an obvious vgaHW locking bug. The rest are minor cosmetics and documentation updates. - Add credit to GB18030 patches to Yu Shao and James Su, whom both worked on this support over time.- Updated CVS snapshot to 4.2.99.902-20030223 - CVS checkin 946 fixes GL debugging junk printed to screen (#84347) - More CVS fixes for Radeon ARGB cursor code from Keith and Michel Daenzer (#84691) - CVS checkin 939 Check pScrn->vtSema before calling xf86SetCursor() from xf86CursorCloseScreen(). This avoids a segfault at exit with some drivers, and I believe should fix the problem reported in (#79678,84689) - CVS checkin 938 fix adding FP native mode for Radeon (Hui Yu@ATI, Kevin Martin). - This should fix problems people have reported on various Radeon Mobility laptops of being unable to select video modes, etc. I can not get into bugzilla to get the bug IDs currently. - CVS checkin 937 - Initialize I2C when primary head has an invalid DDC type for the Radeon driver (Hui Yu@ATI). This should fix some of the DDC probe related problems some users experience during installation which cause the installer to go bonkers.- Bump kernel-drm requirement to 4.3.0- Implemented XFree86-4.2.99.902-ati-radeon-dpms-on-dvi.patch for DPMS on LCD/DFP panels on Radeon driver (#80629)- Removed Xft/Xft-devel goofup from bad specfile diff application - Added XFree86-4.2.99.902-savage-revert-vbe-change-from-X-cvs.patch to back out a VBE change done in XFree86 CVS sometime in the last couple of months, which I believe causes the Savage MX probs (#72476,80278,80346,80423,82394)- Updated CVS snapshot to 4.2.99.902-20030220 - Removed integraged now: XFree86-4.2.99.901-debian-dont_vtswitch_option.patch, XFree86-4.2.99.901-017_neomagic_manpage_fixes.patch, XFree86-4.2.99.901-018_fontenc_null_ptr_deref_fix.patch, XFree86-4.2.99.901-020_proxymgr_manpage_fixes.patch, XFree86-4.2.99.901-035_xim_memleak_fix.patch, XFree86-4.2.99.902-ati-radeon-buildfix.patch- Ok, no kidding this time. I really really mean it, the new savage driver is definitely in this build for sure for sure. No kidding around this time. - Created and added XFree86-4.2.99.902-savage-Imakefile-vbe-fixup.patch, XFree86-4.2.99.902-savage-1.1.26cvs-1.1.27t-fixups.patch, XFree86-4.2.99.902-savage-1.1.26cvs-1.1.27t-accel-fixup.patch - Added XFree86-4.2.99.902-redhat-version-change-notification.patch which does not actually do anything except fail to apply cleanly when the XFree86 version changes, thus alerting me to change the spec file Version field, and update the patch- Actually ENABLE the new savage driver with rpm spec conditional, since I moronically did not enable it the first time- Implemented a new radeon driver config file option "ForceLegacyCRT" in XFree86-4.2.99.902-ati-radeon-alan-legacy-vga.patch to allow Alan to use an ancient VGA mono monitor, and documented it in the radeon manpage (#69976)- Updated CVS snapshot to 4.2.99.902-20030218 - Added Red Hat buildsystem BuildConflicts line to test for disk space - Added Red Hat buildsystem BuildProvides line to hint that up to 8 CPUs can be used effectively when compiling XFree86 - Added XFree86-4.2.99.902-ati-radeon-buildfix.patch to fix ATI Radeon driver build problem- Update savage driver to Tim Roberts tried and true faithful savage driver new release 1.1.27t (#72476,80278,80346,80423,82394)- Updated CVS snapshot to 4.2.99.902-20030217 - Release Candidate 2 - Dropped XFree86-4.2.99.901-xlib-hang.patch, XFree86-4.2.99.901-argb-cursor-render-bugfix-keithp.patch, XFree86-4.2.99.901-xft-_XftSmoothGlyphMono-endless-loop-fix.patch, XFree86-4.2.99.901-xft-crash-with-non-render-server-from-keithp.patch- Added XFree86-4.2.99.901-argb-cursor-render-bugfix-keithp.patch, yet another ARGB cursor bugfix. This one is in the render code. It should fix problems reported by msw and others of random blobs and stuff following the cursors around, in particular when using the 'core' cursor theme - Added XFree86-4.2.99.901-xft-_XftSmoothGlyphMono-endless-loop-fix.patch patch to fix endless loop in libXft Nalin found while investigating (#80414) - Added XFree86-4.2.99.901-xft-crash-with-non-render-server-from-keithp.patch to fix bug http://bugzilla.mozilla.org/show_bug.cgi?id=175108 reported to me by blizzard that crashes mozilla, which is also found in (#80414)- Updated CVS snapshot to 4.2.99.901-20030213, fixing several critical bugs - Disabled keiths argb radeon cursor patch as CVS has some other fix now - New CVS also fixes the same problem on r128 and mach64 mouse cursor that the radeon had (#83284,83768) - Disabled revert-xkb-patch-865 as the problem with VTswitching is fixed now in this snapshot (#84306)- Updated CVS snapshot to 4.2.99.901-20030212- Added XFree86-4.2.99.901-dont-install-Xcms.txt.patch to not install Xcms.txt by default, under recommendation of Keith Packard - Added Conflicts: Xft to -libs, and Conflicts: Xft-devel to -devel, and removed <= 2.0 versioning off Obsoletes: Xft and Xft-devel to try and fix problem where if you upgrade an 8.0 system to 4.2.99.x the Xft2 library becomes unusable (#77838) - Created XFree86-4.2.99.901-revert-xkb-patch-865.patch to revert the changes from Ivan Pascal that Daved Dawes checked into CVS in XFree86 CHANGELOG entry 865 to fix autorepeat on keypad /*-+ and fix mouse-keys, and other things, which had the result of totally breaking VT switching from terminal to terminal after switching out of X- Added XFree86-4.2.99.901-ati-radeon-argb-cursor-fix-from-keithp.patch from Keith, and disabled argb cursor patch for radeon (#83768)- Added XFree86-4.2.99.901-debian-dont_vtswitch_option.patch which adds a new ServerFlags option "DontVTSwitch", to allow disabling of VT switching via CTRL-ALT-Fn written by Branden Robinson for Debian - Added XFree86-4.2.99.901-ati-radeon-ia64-pagesize.patch which removes the hardcoded compile-time pagesize in the radeon and r128 drivers, replacing it with runtime detection with getpagesize() since platforms like ia64, alpha and others that can have different runtime page sizes will work - Added XFree86-4.2.99.901-xlib-hang.patch that fixes a bug in _XEventsQueued() that causes an 'event reader lock' of another thread to be released which can cause a hang and keyboard lockup in KDE. For example, using KDE CVS and xine CVS, by playing a movie clip with Noatun (hang and keyboard lockup after playback has stopped). - Update s3virge driver documentation for 4.3.0 - Added XFree86-4.2.99.901-007_fix_xfree86_man_version_string.diff to fix up XFree86 version string in manpages - Added XFree86-4.2.99.901-008_doc_extensions_fix.diff to sanitize filename extensions of XFree86 documentation for consistency - Added XFree86-4.2.99.901-012_r128_xv_btyeswap_fix.patch patch to fix Xvideo extention support for YUV2 and UYVY in r128 driver on big-endian processors - Neomagic manpage update XFree86-4.2.99.901-017_neomagic_manpage_fixes.patch - Added XFree86-4.2.99.901-018_fontenc_null_ptr_deref_fix.patch which fixes a null pointer dereference in the fontenc library - Added XFree86-4.2.99.901-024_r128_and_radeon_vgahw_independence.patch to fix problem on some architectures that do not allow legacy VGA hardware access - Added XFree86-4.2.99.901-027_ati_driver_message_cleanups.patch to fix up the formatting of a few ATI driver startup messages - Added XFree86-4.2.99.901-028_fbdev_depth24_support.patch to fix fbdev 24bit depth support - Added XFree86-4.2.99.901-035_xim_memleak_fix.patch to fix a memory leak in the XIM code- Updated CVS snapshot to 4.2.99.901-20030211- Moved TTF and OTF subpackage fonts.* and other auxilliary ghost files into the base-fonts subpackage so the dirs are always present, even if the truetype and/or syriac-fonts packages are not installed, and to prevent multiple packages from claiming ownership of the same files (#83913) - Added Conflicts lines for the above change - Added Buildrequires: ed for fontconfig - Removed unapplied ancient XFree86-4.1-xinerama.patch workaround - Added XFree86-4.2.99.901-xauth-diskfull.patch which fixes a bug where xauth may write an incomplete .xauth file and delete the old one if there is insufficient disk space (Harald Hoyer - #84036)- Added more with_foo buildtime conditionals for standardization - Cleaned up post buildroot-install time file deletions of unpackaged files- Updated CVS snapshot to 4.2.99.901-20030209 - Added XFree86-4.2.99.901-xkb-sysreq.patch to fix GNOME printscreen issue due to bad mapping of sysreq (#80373) - Removed XFree86-4.2.99.4-ati-r128-full-chipset-support.patch as it got integrated into upstream sources - Added XFree86-4.2.99.901-ati-r128-chip-names-touchup.patch to make R128 chips that we are not sure are AGP or not show up as "(AGP?)", to heighten the chance of a user noticing it and providing us with this information. - Removed XFree86-4.2.99.4-ati-xf86PciInfo-chip-support-update.patch as it is integrated upstream now- Updated CVS snapshot to 4.2.99.901-20030206 - New Intel i865 integrated video support is in this snapshot - Made XFree86-4.2.99.901-ati-radeon-missing-symbol-drmScatterGatherAlloc.patch to fix a bug reported on devel@xfree86.org, also submitted it- Updated CVS snapshot to 4.2.99.901-20030205 - Removed integrated patches XFree86-4.2.99.3-encodings-Imakefile-fix.patch - Add XFree86-4.2.99.901-ati-radeon-fix-broken-argb-cursors.patch to hopefully fix all the Radeon Xcursor related weirdness (#78231,80445)- Cleanups of specfile, host.def - Ported parallelmake patch to current tree, and experimented with it but could not get it working properly so it is disabled again- Added XFree86-4.2.99.901-chips-default-to-swcursor-on-65550.patch to default to swcursor on C&T 65550 for bug (#82438) - Improved chips-default-to-noaccel-on-69000 patch to be more correct for (#74841)- Updated CVS snapshot to 4.2.99.901-20030203 - Added BuildBethMarduthoFonts YES as XFree86 CVS somehow weirdly stopped building Syriac fonts - Disable mga-xinerama-G450-issue fix as it is believed to be fixed in CVS now with CHANGELOG log entry 818. Needs confirmation to be sure.- Updated XFree86-4.2.99.4-ati-r128-full-chipset-support.2.patch to remove stupid ANSI trigraphs and escape them properly.- Wrote XFree86-4.2.99.4-ati-r128-full-chipset-support.patch to add 2D and 3D support for 30 previously unsupported ATI Rage 128 chips including various Rage 128 Pro, Rage 128 Pro Ultra chips- Wrote XFree86-4.2.99.4-ati-xf86PciInfo-chip-support-update.patch to update the PCI ID list for further driver update patches to follow - Wrote XFree86-4.2.99.4-ati-radeon-8500-9100-firegl-8700-8800-update.patch to add 2D and 3D support for 8 previously unsupported chips in the ATI R200 chip family including various Radeon 8500, FireGL 8700/8800, and the new Powered by ATI Radeon 9100 from Sapphire - Added XFree86-4.2.99.4-ati-radeon-TV_DAC_CNTL.patch minor code cleanup- Added XFree86-4.2.99.4-locale-alias-openi18n.patch from Leon Ho, required for openi18n compliance - Cleaned up post build/install file/dir deletions to move gradually towards being able to turn on _unpackaged_files_terminate_build- Removed handhelds cursor theme - Added XFree86-4.2.99.3-ati-chip-support-update.patch, partial support patch for ATI Rage 128, Radeon, etc. hardware previously unsupported- Update CVS snapshot to 20030129 - Removed patches present upstream now: XFree86-4.2.99.4-xkb-se-fix.patch XFree86-4.2.99.4-xkb-rules-xfree86-ca_enhanced.patch - Updated XFree86-4.2.99.4-redhat-custom-startup.patch to work with latest- Removed from host.def HasMMXSupport, Has3DNowSupport, HasKatmaiSupport, and HasSSESupport after manually verifying the current default Imake config files enable this properly on x86 and x86_64 - Removed -fPIC from global x86_64 flags again, and force StaticNeedsPicForShared to YES on all architectures in host.def, instead of just on x86 - Added XFree86-4.2.99.4-IncludeSharedObjectInNormalLib.patch to fix up remaining issues with StaticNeedsPicForShared (#78188,82954) - Added XFree86-4.2.99.4-x86_64-glx-nopic.patch to build GLX nopic on x86_64 for performance reasons - as it is on x86- Rebuilt with -fPIC temporarily for anaconda et al. on hammer- Rebuilt without -fPIC on hammer for debugging X server- Added more complete XFree86-4.2.99.4xkb-se-fix.patch from David Dawes to replace XFree86-4.2.99.4-xkb-se-asciitilde.patch if it works properly - Obsolete XFree86-Setup (#82674) - Added XFree86-4.2.99.4-xkb-rules-xfree86-ca_enhanced.patch to fix (#80425)- Added XFree86-4.2.99.4-xkb-se-asciitilde.patch to fix (#80922,82097)- Update CVS snapshot to 20030121 - Drop patches now included upstream ati-radeon-mono-8x8-accel-fix, mga-vbe-dvi-ddc, xmag fix - Added missing Conflicts: XFree86-sdk < 4.2.99.3-20030115.0- Flagged {_x11icondir}/defaults/index.theme file as config noreplace - Disable building on ia64 due to compiler SEGV- Bump kernel-drm req up to 4.2.99.3 to match new kernel DRM - Remove -fPIC from x86_64 default opt flags, and disable ELF loader patch - Add XFree86-4.2.99.3-Imake-make-icondir-configurable.patch to make XFree86 Xcursor ICONDIR and XCURSORPATH build time configurable and default the cursors to install into /usr/share/icons - Added new specfile directory macro _x11icondir and use it everywhere - Made default cursor theme Bluecurve- Added XFree86-4.2.99.3-ati-radeon-mono-8x8-accel-fix.patch to fix mono 8x8 color expansion issue causing mirroring on RV200+- Update CVS snapshot to 20030118 - Add XFree86-4.2.99.3-xmag-segv-uninitialized-pointer.patch to fix (#81966) - Major radeon driver updates and fixes: Fix for Mono8x8 patterns on Radeon, Fix for Radeon mode validation, Workaround for flickering problem with switching between ARGB and mono cursors on Radeons, DDCMode fix for VidMode extension, Panel detection bug fix for Radeon, Add Xv overlay support for dual headed Radeons, Fix Radeon driver 24-bit support for flat panels, Add Radeon 9500/Pro support, Disable CGWorkaround for non-A11 rev R300s, Radeon solid/dashed line fix for RV200 and newer cards, Radeon overlay gamma fix, Radeon LG panel fix. - Add XFree86-4.2.99.3-elf-loader-x86_64-relocation-fixups.patch and disabled dllmodules to try and fix ELF relocation problem reported in bug (#81984)- Add XFree86-4.2.99.3-mga-vbe-dvi-ddc.patch to support for DDC probe of DVI panels on Matrox hardware - Use MakeDllModules on x86_64 as ELF loader is currently problematic- Update CVS snapshot to 20030115 - Move syriac fonts from TTF subdir to new OTF subdir - Move inb inl inw ioport outb outl outw commands to XFree86 pkg from sdk- Added XFree86-4.2.99.3-mga-unresolved-symbols.patch to fix unresolved DRM symbols in mga driver for bug (#80968)- Added fontconfig-slighthint patch from our fontconfig package as a hack to allow the embeddedbitmap and xft-loadtarget patches to build correctly. - Added XFree86-4.2.99.3-embeddedbitmap.patch - Added XFree86-4.2.99.3-encodings-Imakefile-fix.patch to fix (#81565)- Add Requires: xinitrc (#81424) - Remove placeholder redglass cursor theme - Disable accel on C&T 69000 XFree86-4.2.99.3-chips-noaccel.patch for (#74841)- Disable fc-cache in post/postun scripts except for TTF/Type1 directories. - Created XFree86-4.2.99.3-savage-pci-id-fixes.patch to fix up the Savage driver PCI IDs that were missing and/or incorrect.- Disabled the XFree86-4.2.99.3-xft-loadtarget.patch patch as it is broken.- Remove xman as there are other manpage viewers, and this one seems to be unmaintained upstream. Conditionalized with with_xman (#81003) - Update dependancy on freetype to version 2.1.3-4 which contains a patch that a patch we are applying to Xft requires.- Added XFree86-4.2.99.3-xft-loadtarget.patch from Owen Taylor, which is a patch needed for our freetype 2.1.3 package, ported forward from the 8.0 Xft package- Updated CVS snapshot to version 4.2.99.3 20021230 - Added XFree86-4.2.99.2-netmouse.patch to fix mouse protocol initialization problem reported on XFree86 bug report list - Removed the gtf utility, its manpage, and libxf86config.a from s390/s390x - Removed leading "0." from specfile RELEASE field as it is unnecessary- Disabled and removed whiteglass cursor theme as it is broken and nobody is likely to fix it it seems. We will be supplying our own theme in the end that does not suck (#75844)- Updated CVS snapshot to 20021217, dropped unneeded patches - Added XFree86-4.2.99.2-xinerama-G450-issue.patch to work around mga issues when RENDER is used with Xinerama due to faulty Render acceleration- Updated CVS snapshot to 20021215, dropped unneeded patches - Update BuildRequires freetype to >= 2.1.2-7 - Added ISO8859-14 fonts for Welsh for Alan - Disable XFree86-sdk by default for now due to nsc driver sdk issue - define debug_package to disable -debuginfo packages for now- Added Buildrequires: libpng-devel for bug (#79623)- Updated CVS snapshot to 20021210, dropped unneeded patches - Added XFree86-4.2.99.2-mga-unresolved-symbols.patch - Added /usr/X11R6/lib/X11/Xcms.txt to libs-data - Changed XFree86-devel requires from fontconfig to fontconfig-devel (#79183) - Added alpha-domain and alpha-domain2 patches back (split original in 2, and updated)- Updated CVS snapshot to 20021209, dropped unneeded patches - Numerous bug fixes, stability, and correctness fixes for the Intel 830/845G 2D, 3D, and Xvideo support - Updated Korean font encoding tables XFree86-4.2.99.2-korean-encoding-table-fixes.patch - Added XFree86-4.2.99.2-20021209-cvs-update.patch to further update to latest CVS as David checked in a lot more stuff - Removed patches included upstream now: XFree86-4.2.99.2-more-lib64-fixes.patch XFree86-4.2.99.2-disable-DRI-on-s390-by-default.patch, XFree86-4.2.99.2-x86_64-has-SSE-MMX-3DNOW.patch- Updated XFree86-4.2.99.2-redhat-custom-startup.patch to handle bug (#77930) - Added XFree86-4.2.99.2-alpha-domain.patch to fix PCI domains on alpha, and re-enabled building on alpha. - Removed with_ttmkfdir conditional, all ttmkfdir sources/patches and building- Updated CVS snapshot to 20021208, dropped unneeded patches- Fixed XFree86-4.2.99.2-mkfontdir-perms.patch to correctly chmod 0644 (#79119) - Added XFree86-4.2.99.2-xkb-gb-backslash-and-bar.patch to fix missing backslash/vertical bar keys on gb keyboard- Enabled StaticNeedsPicForShared YES in host.def for x86. xfree86.cf defaults to YES for all archs except x86, which causes problems linking X .a only libs into KDE/GNOME .so libs and other problems like libxkbfile.a in (#78188) - Disabled older temporary hack XFree86-4.1-xinerama.patch to see if it is needed- Get rid of xtrap-clients and cursors subpackages, incorporating them into main XFree86 package and adding an Obsoletes tag for them - Added missing syriac-fonts post/postun scripts - Added XFree86-4.2.99.2-omGeneric-stop-app-crashes.patch to fix problems with apps crashing, due to some bad juju recently checked into upstream CVS- REALLY fix VT switch this time by making patch20007 apply as 20007 and not as 20006 doofus- Added XFree86-4.2.99.2-TEMPORARY-fix-for-broken-vt-switching.patch to fix broken VT switching, CTRL-ALT-BS, etc...- Re-enabled the old manual stripping policy and disabled rpm stripping as it has shown to not strip everything that needs to be stripped. - Added -fPIC to global compiler options for x86_64, as static libs need to be built with -fPIC to be linked into shared libs- Added XFree86-4.2.99.2-fix-CppCmd-path.patch to change the location X looks for cpp from /lib/cpp to /usr/bin/cpp- Added "-fno-strength-reduce GccAliasingArgs" to x86 and x86_64 default optimization flags, and "GccAliasingArgs" to default alpha and ppc flags, since we override XFree86 defaults, and they include these optimizations.- Add XFree86-4.2.99.2-missing-carriage-return-before-endif.patch to fix Alpha build failure in axpPci.c due to one backspace too many... - XFree86-4.2.99.2-build-fixes-for-alpha.patch- Woohoo! We now let rpm strip the XFree86 build as it no longer fries X modules - Added "BuildRequires: rpm >= 4.0.4" for safe rpm stripping, as older versions of rpm will strip XFree86 modules and make them unloadable - Added new _x11localedir macro, did search and replace cleanup - _includedir cleanup - Got rid of the Imake defines (BuildXinerama, BuildXineramaLibrary, BuildXF86DGA, BuildXF86DRI) from host.def, as they should be set properly in stock XFree86 configs, and if not, patched and sent upstream - Remove ancient BuildTdfx option that has not been used/needed for ages- Update CVS snapshot to 20021126 - Removed patches that are now upstream: XFree86-4.2.99.2-redhat-xdm-pam.patch XFree86-4.2.0-xkb-french-canadian.patch, XFree86-4.2.0-xkb-is.patch, XFree86-4.2.0-manpage-document-nomtrr-option.patch, XFree86-4.2.0-im-clientmessage-crash.patch- Test build- Test brp-compress out to see if we need to override it still- Added XFree86-4.2.99.2-neomagic-2160-disable-broken-accel-for-bug55587.patch to try to resolve bug (#55587) - Added XFree86-4.2.99.2-disable-DRI-on-s390-by-default.patch to disable DRI on s390 and s390x by default as X was defaulting to DRI=YES and s390 has no video hardware. - Added XFree86-4.2.99.2-x86_64-has-SSE-MMX-3DNOW.patch to enable SSE/MMX/3DNOW on x86_64 by default - Added XFree86-4.2.99.2-more-lib64-fixes.patch to cleanup Imake lib64ness - Modified strip section to call strip once for strip-debug, and once for strip -R .comment, as apparently calling them together does not do what I expected it to do. Thanks to Michael Fratoni for pointing this out.- Reordered more patches to sections 10000/20000, and minor housekeeping - Removed BuildRequires on pam header file, and replaced with pam-devel - Removed XFree86-4.0.3-xaw-asciisink.patch, as a different solution was put into upstream sources- Updated CVS snapshot to 20021122 - Updated XFree86-4.2.99.2-redhat-custom-modelines.patch as extramodes changed - Removed Alan Cox's input drivers, as they're included upstream now - Removed patches that are now included upstream: XFree86-4.2.0-minor-typo.patch, XFree86-4.2.0-hostname-stderr-fix.patch, XFree86-4.2.99.2-rename-Katmai-to-SSE.patch XFree86-4.2.99.2-new-alan-cox-input-drivers.patch, XFree86-4.2.99.2-Imake-lib64-fixups.patch XFree86-4.2.99.2-ati-r128-ppc-dri-buildfix.patch, XFree86-4.2.0-bad-mmap-check.patch XFree86-4.2.99.2-FontconfigFontsConfDir.patch, XFree86-4.2.99.2-Imakefile-fontconfig-build-dir-fixup.patch - Removed XFree86-4.1.0-ppc-compiler-workaround.patch, we are building with gcc 3.2 now, so this hack for older gcc 2.96 on ppc is not needed anymore - Submitted many of the remaining patches upstream in hopes they will be applied soon, and we can drop many more of them - Re-ordered the patch section a bit, removed dead patches that were still being included but not applied. Made 2 new sections for patches - range 10000 is for Red Hat custom patches that are specific to Red Hat Linux, and which are not particularly relevant for upstream or other distros, and a second section for patches which are certified ugly hacks, but needed, and possibly refused by upstream (with commented reason hopefully). - Added new gtf utility and manpage to main package for calculating new VESA GTF (General Timing Formula) modelines for creating new video modes- Added BuildRequires: expat-devel (#78422) - Removed s390 FreeType patch, as Than says it is no longer needed.- Conditionalized with_sdk to exclude it on s390 - Do not build/include mmapr and mmapw on s390 and ppc platforms - Conditionalized inclusion of libxrx and disabled it as it is a Netscape plugin apparently, and a potentially serious security risk, and upstream recommends not to ship it.- Disable ati-r128-vtswitch-fix-busmaster-enable patch to see if the problem still exists or not. - Explicitly list each .a and .so library in the -devel package, so that unwanted libs that show up due to XFree86 Imake bugs do not get packaged. If RPM is configured with _unpackaged_files_terminate_build == 1, these types of errors can get picked up and fixed before becoming a problem. - Updated fontconfig-devel BuildRequires dependancy to version 2.1- Disabled xf86config, as it is ancient, and unneeded. Our supported config tool is redhat-config-xfree86 now, and we want people using it, testing it, reporting bugs and feature requests to us in bugzilla. - Re-enabled input drivers from Alan Cox - Moved pkgconfig files, from _x11libdir to _libdir (#77815,77858) - Changed host.def to use HasFontconfig instead of UseFontconfig - Added XFree86-4.2.99.2-Imakefile-fontconfig-build-dir-fixup.patch to fix a bug in xc/lib/Imakefile which caused "make clean" to fail during build time when HasFontconfig is defined. The problem is due to the FontconfigLibDir being added to $SUBDIRS always, and all targets seeing it and barfing when it is /usr/lib. I changed it to FONTCONFIGBUILDDIR instead, which only gets defined when HasFontconfig is not defined. - Changed default cursor theme to whiteglass so that more people include with graphic art abilities see the graphic glitches and maybe one of them will fix the broken whiteglass theme - Added XFree86-4.2.99.2-FontconfigFontsConfDir.patch to rename conflicting namespace in Imakefile - Re-enabled the Imake-lib64-fixups patch for proper building on x86_64- Enabled kernel DRM check to ensure user has a Red Hat rawhide kernel installed which supplies the version of DRM that is required for proper operation of DRI. - Added rename-Katmai-to-SSE patch to rename the HasKatmaiSupport Imake define to be more correct as to what it is really doing.- Added XFree86-4.1.0-ia64-vgaHW-memory-barrier2.patch which adds memory barriers to the vgaHW functions to guarantee memory ordering. This resolves MCAs on ia64 and possibly other platforms - Added memory barriers to nv driver in nv-ia64-memory-barrier.patch to solve HP Itanium 2 machine checks- Added patch to fix Imake related lib64 issue on Hammer where any library chosen with HasSomelib results in Imake looking in /usr/lib rather than in /usr/LibDirName - XFree86-4.2.99.2-Imake-lib64-fixups.patch- Removed glFrustum-fixie patch as it is unneeded in CVS X - Removed various other patches that are no longer needed, were not being applied, but were still being included in spec file - Updated redhat-custom-startup.patch to apply cleanly to 4.2.99.2 - Enabled HasFontconfig in host.def (duh duh duh)- Specfile cleanups. Removed stuff custom to my local environment that is now unneeded, and fixed some build issues.- Simplified the XFree86-75dpi-fonts and XFree86-100dpi-fonts file lists by using %exclude globs instead of individually listing a billion oddball globs, in order to shrink specfile, and improve maintainability. - Reordered the files lists to be closer to package alphabetical order- Added new __fccache macro for running fc-cache if it exists when installing or uninstalling font packages. Added __fccache to all font packages post/postun scripts. (#77542) - Conditionalized the new inb, inl, etc. commands to only be included on architectures that they build and work for.- Updated CVS snapshot to Nov 10th - Added new inb, inl, inw, ioport, outb, outl, outw commands to SDK subpackage as they are potentially useful for driver development on architectures that they work on. - Added ownership of _x11includedir dir to XFree86-devel package- Updated various Requires lines in subpackages that specified a dep on another XFree86 subpackage to require the same ver-rel instead of just ver. - Reorganized font subpackages %description near the %package definition - Made font subpackage inclusion conditional with the macro with_fonts with hopes to build the fonts in a separate noarch src.rpm in the future instead of being part of the XFree86 build process. (untested yet) - Fixed bad Conflicts line missing a <= in XFree86-libs-data- Enabled the restest Xresource client - Disabled experimental building of kdrive/TinyX - Created new XFree86-libs-data subpackage to hold architecture independent data files including locale/compose datafiles, XErrorsDb, rgb.txt and other files that should be installed with the X libraries, and not require the main XFree86 package to be installed.- Updated CVS snapshot to Nov 5th - Includes bug fix for i830/i845 hardware/software cursor issue (#76772,77213)- Changed all occurances of /usr/X11R6/lib in all pre/post/postun/trigger/etc. scripts to use either _x11libdir or _x11datadir, whichever is appropriate, so that things work properly on x86_64 and other lib64 archs - Replaced all occurances of {_x11datadir}/X11/fonts with {_x11fontdir}- Removed Xterm3D from specfile. There are enough terminal emulators in the distribution already. xterm3D causes build dep loops, and is something that is more suited to its own RPM package IMHO. - Added conditionalized Obsoletes: XFree86-xf86cfg - Added XFree86-4.2.0-bad-mmap-check.patch to fix (#71783)- Changed SDK to use a filelist instead, and added code to %install to generate the filelist.sdk. Hopefully this will work on all arches now. - Added XFree86-4.2.99.2-ati-r128-ppc-dri-buildfix.patch to fix build problem on PPC when DRI is not being built in.- Updated XFree86-sdk package to use _x11libdir instead of _x11datadir as the sdk gets installed to /usr/lib which is lib64 on Hammer.- Updated CVS snapshot, includes DRI CVS merge with Radeon 8500 3D support, various RandR and Xcursor fixes. Note: You need to compile DRM by hand from XFree86 CVS for DRI to work, and no I wont help, I am busy. Mailing lists exist for people who need help compiling DRM kernel modules. - Cleaned up cursors subpackage and added handheld cursors- Added files to sdk filelist. Probably a lot of them can be commented out as they dont really suit the purpose the subpackage is being added, but I have included them anyway for now.- Updated CVS snapshot to 4.2.99.2- Added new XFree86-sdk subpackage in order to allow X driver modules to be compiled in separate subpackages, and to facilitate driver development outside of a fully installed XFree86 source code tree. - Fixed broken geode driver Imakefile SDK installation code- Disabled kdrive on ia64 also, as it fails to build (dunno why because the buildsystem deleted the log). No biggie as I doubt we will see ia64 based PDAs anytime soon.- Completed the with_kdrive conditional build stuff. TinyX kdrive etc. X servers can now be optionally enabled to be built. WARNING: Note that this is not something which will be officially included or supported in Red Hat Linux, but rather it is intended solely for my own experimentation with kdrive. Do not use this thinking or expecting that it will be enabled in official packages. No guarantees, you have been warned. Have fun with it for now, while I have it enabled anyway. - Disabled the kdrive option on x86_64 and ppc because gcc exhibits an ICE when building XIgsServer on x86_64, and the build fails on XIgsServer due to a source bug it appears when building on ppc - Added XFree86-4.2.99.1-kdrive-posix-sigaction.patch to fix sigaction issue in kdrive - sa_restore is obsolete and should not be used, POSIX does not require it, and it causes the build to fail on alpha. - Disable kdrive build on alpha, due to build failures- Updated file lists to conditionally include all freetype2 related files in the appropriate subpackages when with_freetype2 is defined- Updated CVS snapshot to the latest bits - Removed patches that are included in CVS code now: XFree86-4.2.1-xshm-header-file-fix.patch, XFree86-4.2.99.1-drm-device-dir-perms.patch- Added Provides/Obsoletes for Xft/Xft-devel for obsolesense and compatibility- Added new manpages for bdftruncate, cleanlinks, mergelib, mkhtmlindex, ucs2any - Added mmapw binary- Excluded fontconfig development files in the XFree86-devel subpackage when with_fontconfig is not defined, by using the rare rpm 'exclude' directive, both so that it does not conflict with existing fontconfig libs, and also to make people wonder what the exclude directive is and cause confusion and world unrest, just because.- Added xc/programs/Xserver/hw/xfree86/CHANGELOG doc flagged to main XFree86 subpackage, so people can find the developmental CHANGELOG without installing the source code- Updated CVS snapshot to 2002-10-14, which includes xfs updates and cleanups, APM suspend fixes, Radeon crash freeze fix, RandR extension updated to support refresh rate changes. - Made XFree86-libs subpackage Obsolete: Xft (#75932) - Fixed XFree86CVSDate in spec file to be correct. Doh! - Disabled xkb-fix-deadkey-quotedbl-us_intl patch as one hunk fails, and it looks like CVS includes these changes already. I really wish people would file these type of change requests upstream FIRST so I do not have to maintain separate fork of them. - Removed XFree86-4.2.0-xfs-syslog.patch, XFree86-4.2.0-sane-deadacc.patch, XFree86-4.2.1-lbxproxy-SEGV-fix.patch, XFree86-4.2.1-vidmode-manpage-fixes.diff, as they are included in CVS now. - Updated drm-device-dir-perms.patch, xtt.patch, Wacom patch to account for changes in CVS update, and cleaned up patches a bit. - Wrapped %clean section of specfile with auto_mharris_mode check, so that the BUILD_ROOT files get left behind for me for doing rpmfilediff to check for new files. - Removed .pl filename extensions from ucs2bdf.pl and bdftruncate.pl in file lists, to match the newly generated filenames.- Updated CVS snapshot to 20021012 - Added changelog entries below for CVS updates. ;o) - Added /usr/X11R6/bin/xft-config to -devel pkg - Made -libs package own the dirs _x11libdir, and _x11datadir - Fixes and improvements for the en_US.UTF8 compose rules, and for some dead accents for iso8859-1 and iso8859-15 (Alexandre Oliva). - Fix 24bpp displays with the fbdev driver when not using shadowfb - New "kbd" keyboard driver module which will eventually replace the built in "keyboard" driver. - New Natsemi geode video driver from Alan Hourihane- Made specfile changes for x86_64 arch (AMD Hammer) for lib64 support, which involved adding a new macro _x11datadir and making all fonts, and other non executable data, etc. use the _x11datadir macro instead of _x11libdir. On x86/alpha/ia64/ppc _x11datadir and _x11libdir are the same, however on x86_64 architecture, _x11libdir=/usr/X11R6/lib64 and _x11datadir=/usr/X11R6/lib - The package currently builds on x86/x86_64/alpha/ia64/ppc. ppc64/s390/s390x to be tested soon, as will sparc/sparc64 if Tom Callaway gives it a shot.- Updated file lists for new files in CVS. Normally I would add details as to what new applications were added, but I am feeling lazy. Basically, mmapr, new xcursor stuffs, and a few other things. Blah blah blah. - Created new subpackage XFree86-cursors for the new truecolor antialiased, alpha blended, animated, themeable mouse pointers in XFree86. Go Keith go! - Created new subpackage XFree86-syriac-fonts for the new Syriac fonts. - Made various updates to _x11foodir macro usage throughout the specfile- Forked XFree86.spec and updated XFree86 to CVS snapshot from head of CVS dated 20021008 - Removed all patches from specfile which are now present in CVS, or otherwise made irrelevant. There are still various patches disabled which are not yet known if they are needed or not, and might need porting work.- Updated Patch0 to current xf-4_2-branch, picking up a few security fixes and bugfixes, in particular for MIT-SHM, xdm - Removed XFree86-4.0.3-xaw-freed-mem-deref.patch now as it _finally_ is in XFree86.org sources. ;o) - Removed XFree86-4.2.0-xdm-pam-conv-and-realloc.patch - it is integrated now- Various architectural cleanups for all-arch build- Added new input drivers from Alan Cox for "Fujitsu stylistic", Palmax PD1000/PD1100 Input driver, Union Reality UR-98 head tracker- Added new Cyrix driver 0.2 from Alan Cox, which fixes (#72297)- Updated savage driver to the upstream maintainer Tim Roberts new 1.1.25t version which solves various savage driver problems, which are fully documented at http://www.probo.com/timr/savage40.html (#71973)- Disabled DRI on ppc at streeters request- Added XFree86-4.2.1-lbxproxy-SEGV-fix.patch from Debian to fix a SEGV in lbxproxy - Added XFree86-4.2.1-xshm-header-file-fix.patch from Debian - Added XFree86-4.2.1-vidmode-manpage-fixes.diff from Debian- Added ati-r128-dri-dga and ati-radeon-dri-dga patches to fix bugs in the Radeon and Rage 128 drivers which cause crashes while using VMware with DGA when DRI is enabled (#73678) - Fix up ttmkfdir to look for freetype in _libdir for AMD Hammer- Remove scanpci, pcitweak from ppc/ppc64 builds- (x86_64) Added Imake define to point Freetype2LibDir to _libdir- Replaced the BuildRequires on Glide3 and Glide3-devel that was conditionalized for x86/alpha/ia64 with some new shell script code in the prep section to ensure Glide3 is installed on the build system since some rpm based software freaks out with conditional BuildRequires, and I am a nice guy.- Did a search and replace on specfile to convert all locations of the X11 font directory to _x11fontdir, /usr/X11R6/bin -> _x11bindir, and similar for _x11mandir, etc. throughout the specfile- Updated XFree86 to version 4.2.1. This does not bring us anything new since 4.2.1 is merely 4.2.0 stable branch (which our 4.2.0 was already tracking all along) plus the latest X security patch, which 4.2.0-72 contains. This release is merely a cosmetic update to official 4.2.1 in order to quell all of the "Why doesnt Red Hat have 4.2.1" questions that are sure to abound. - Removed patches present in 4.2.1: XFree86-4.2.0-vesa-xf86SetDpi-fix.patch, XFree86-4.0.3-xaw-readtext.patch, XFree86-4.1.0-time-wraparound.patch, XFree86-4.2.0-mitshm.patch, XFree86-4.2.0-xkb-us_intl-missing-commas.patch, XFree86-4.2.0-overlay-crash.patch- Fix zh_TW-xcinNEW patch which lost a couple of things in the merge with the XLOCALEDIR patch- Added xlib-security-fix-XLOCALEDIR to fix a security vulnerability in Xlib i18n code new to 4.2.0. 4.1.0 and earlier is not susceptible. - XFree86.org has tagged a new release of XFree86 4.2.1 in CVS for this issue. Red Hat current XFree86 4.2.0 is essentially 4.2.1 except for the cosmetic version number difference - Updated the zh_TW-xcin patch to zh_TW-xcinNEW which handles a code area that overlapped with the new security patch - Updated the _LP64 kludge patch on alpha/ia64 as it also overlapped the security patch- Updated custom startup log patch to fix a few glitches - Added -fPIC to x86_64 compiler flags until better solution is made in future - Added XFree86-4.2.0-libXrender-bugfix.patch to fix showstopper (#73243)- Added Obsoletes: XFree86-compat-libs to list of obsolete packages- Rebuild with a few patches disabled that errored out on me and will be fixed in future build- Pruned ia64 driver list based on unknown state of working order for these drivers on ia64 to minimalize support issues. As hardware becomes available for testing, and drivers can be verified to be mostly sane/working, they can be re-enabled on a driver by driver basis. - Added XFree86-4.2.0-im-clientmessage-crash.patch to fix CJK application crashing showstopper bug (#72236)- Actually properly enable the maxxfbmem patch from 4.2.0-67.5 this time - Fix unresolved symbols in SiS driver reported by Alan Cox- Changed buildsys autodetect to use new beehive mechanism (#71055) - Added kernel version string to X server startup messages for easier debugging - Fixed SiS hardware cursor (hopefully) with XFree86-4.2.0-sis-cursor.patch - Fixed SiS maxxfbmem driver comparison inversions- Cleaned up 3.3.6 server Obsoletes lines et al. to be more compact- Integrated PPC specfile changes - Enable DRI on ppc, but not ppc64 - PPC using gcc 2.96 autodetection and compiler flag tweaks, etc. - Added ppc64, s390x, x86_64 to conditionalized LP64 patch - PPC - build pcitweak, scanpci on ppc, ppc64 now- Added XFree86-4.2.0-elf-loader-empty-symbol-table.patch to fix gcc 3.2 related issue in the X ELF loader (#70576) - Re-enabled the Radeon DDA removal patch now that Null beta is released- Due to a kernel bug breaking SuS compliance in the Alpha kernel, a change has been made to the install section so that fonts.dir, encodings.dir files are mode 644 before "touch" touches them. - Modified stripping policy at end of install section to strip all ELF executable files in the buildroot instead of just the ones found in /usr/bin and /usr/X11R6/bin. - Modified shared library stripping policy for all libs found in /usr/X11R6/lib, /usr/lib, and now also /usr/X11R6/lib/X11/locale/common. All .so.* libs are now stripped with "strip --strip-unneeded -R .comment" following the current RPM default for stripping shared libraries, and under the advice of Uli. - Note: XFree86 server modules are not stripped at all currently for safety. Once I determine what can be safely stripped, I will add stripping of the modules also.- Removed XIE and PEX5 documentation from XFree86-doc subpackage because both of these extensions are obsolete and not included anymore. - Temporarily disabled XFree86-4.2.0-ati-radeon-dda-removal.patch as it is untested, and I dont want to destabilize. Will re-enable for testing soon.- Added XFree86-4.2.0-ati-radeon-dda-removal.patch which removes the DDA code from the radeon driver which was leftovers from the r128 driver and apparently causes 1400x1050 and other modes to cause an FPE on some radeon boards in some refresh rates (#63593) - Explicitly delete unshipped files from RPM_BUILD_ROOT at end of install section in order to pass RPMs new unpackaged_files_terminate_build option - Moved around some file removals to the new DELETE section of specfile- Added XFree86-4.2.0-ati-mach64-dsp-rounding-error.patch to fix a rounding error in Mach64 DSP calculation - Added XFree86-4.2.0-ati-radeon-missing-commas.patch which fixes a missing comma issue in a struct in radeon_driver.c - Bumped spec file release from 60.4 to 63 due to someone elses package release number conflicting.- Fixed an xdm SEGV bug in verify.c which could be occasionally triggered, due to malloc() allocated memory not being zeroed out. Also fixed a bad realloc() call that assumed the reallocated block of memory would be in the same location every time - fixes bug (#40729)- Added XFree86-4.2.0-sane-deadacc.patch to fix dead diareisis problem reported by aoliva in bug (#71014)- Disabled the ugly 1980 style ancient UNIX ugly ugly X server startup stipple pattern.- Added XFree86-4.2.0-locale-alias-utf8-euro.patch to workaround bug in fi_FI.UTF-8@euro and pt_PT.UTF-8@euro locales (#67914) - Added XFree86-4.2.0-i810-vtswitch-sync-fix.patch to fix i810 VTswitch lockup problem - BLOCKER bug (#66187,#53231)- Removed /usr/share/fonts/default/TrueType from default xfs config file in order to fix BLOCKER bug (#68126,#70178) - Made X resource extension off by default for official builds, due to irrational ABI compatibility concerns. People wanting to use it can easily rebuild X from rpm after setting the define with_Xresource_extension to 1.- Added restest dumb sample client to use the new Xres extension - Conditionalized spec file inclusion of Xres and the sample client- Bugfix to dixsym.c for the Xres extension from yesterday.- Backported the new X-resource extension from XFree86 head CVS, which allows internal X server resource statistics, etc. to be queried by client applications. This can aide in debugging client related pixmap leakage, and a variety of other various resource related problems/bugs in both client applications, as well as the X server itself.- Disabled ParallelBuild on s390 upon request of pknirsch- Added support to the Radeon driver CP engine to perform accelerated color expansion and image writes. The CP engine is used when DRI is enabled. XFree86-4.2.0-ati-radeon-cp-colorexpansion-imagewrite-enhancement.patch- Removed the following paths from the default xfs config file as we do not ship fonts in these directories by default or they dont belong in the default config: /usr/X11R6/lib/X11/fonts/CID, /usr/X11R6/lib/X11/fonts/local, /usr/X11R6/lib/X11/fonts/latin2/Type1, /usr/share/AbiSuite/fonts (#68126)- Added XFree86-4.2.0-ati-r128-vtswitch-fix-busmaster-enable.patch to fix the exact same bug in the r128 driver as was just fixed in the radeon driver in 4.2.0-56. (#62442,#65136)- Added XFree86-4.2.0-libICE-remove-stupid-delay.patch to remove stupid 5 second delays from libICE (#66751) - Added XFree86-4.2.0-ati-radeon-vtswitch-fix-busmaster-enable.patch to fix lockups reported in XFree86 4.2.0 for a long time now by users of ATI Radeon cards with certain hardware combinations. (#62171,#65330)- Added -f flag to hostname command to attempt to fix Custom build - Added fix for SysRq / Print Screen: XFree86-4.2.0-sysreq.patch (#69743)- Made all instances of "rm" in spec file "rm -f" in order to force removal of the file without prompting. Reported by a few people that rebuilding X caused it to hang while compiling on an "rm" command. (#69713) - Added XFree86-4.2.0-manpage-document-nomtrr-option.patch to document Option "NoMTRR" in the XF86Config manpage.- Added ati-rage-xl-pci-spec-cleanup patch to fix ATI Rage XL PCI spec violations (#69291) Replaces previous patch for IBM/Dell/others from Feb 28, 2002 (#58188)- rebuild using gcc-3.2-0.1- Redirect stderr+stdout of hostname command in xon script to prevent stderr from being displayed ot user (#67323) - Corrected incorrect filepath in XftConfig-README-OBSOLETE file (#68632,#69104)- Previous ia64/alpha fix from Jul 08 removed the Xft1 patch, and made the ia64/alpha patch apply on all architectures. Fixed. - Added XFree86-4.2.0-overlay-crash.patch to fix bug (#68516) - Added patch for module loader from Bill Nottingham for gcc 3.1 on ia64 - Add patch to enable/disable DMA for Xv on r128 (#64503,#68058,#68668)- Fix duplicate application of patch4201 on ia64/alpha - Remove ExcludeArch: alpha ia64- Reordered the spec install section so that Xft1 .so gets deleted after it the code which generates the .so devel symlinks - Fixed libXft.so.1 to point to libXft.so.1.2 instead of 1.1, and made sure the files are explicitly listed in the files section.- Added Obsoletes lines for remaining 3.3.6 servers that were not previously obsoleted from the olden days. - Really made Xft headers excluded this time- Assimilated mkxauth into the XFree86-xauth package, and added Obsoletes and Provides entries for it. - Removed Xft headers when new Xft1 is used, as Xft2 now provides the Xft headers instead.- Replaced libXft with the new Xft1 from XFree86 CVS which is completely compatible, however uses fontconfig instead of XftConfig. This has been conditionalized so that it can be disabled for builds for RHL 7.3 et al. The new libXft is Xft 1.2 whereas the stock Xft is 1.1. The new lib should be binary compatible and source compatible. - xftcache conditionalized as well - replaced with fc-cache in fontconfig- Backported the Trident driver from CVS, picking up many Trident fixes from Alan Hourihane- Backported i810/i830 driver from CVS, adding 2D only i845 support - Removed i830m workaround for Dell c400 laptop as it did not work - Backported siliconmotion driver from CVS, picking up a few bugfixes - Conditionalized spec file using backup prefixes on patches- Backported i740 driver from CVS - i740-driver-update-cvs-20020617 - Renamed the radeon-mobility-LX patch to radeon-mobility-FireGL-7800 after determining the two are one and the same card. - Added mkfontscale from XFree86 CVS so we can begin moving away from ttmkfdir and towards mkfontscale which is going to be better in the long run. - Conditionalized inclusion of XftConfig file and Xft/Xrender libraries, so they can be disabled for builds using fontconfig (which replaces them). - Added new /etc/X11/XftConfig-README-OBSOLETE file to explain wht it is no longer there.- Reorganized patch section, moved all video driver patches into an alphabetically sorted driver patch section, and renumbered all patches to meet the new patch numbering scheme. - Backported cirrus driver from XFree86 CVS (4.2.99.1)- Created XFree86-4.2.0-xman-manconf.patch to fix the xman man config file path for once and for all. Sent upstream so hopefully this will get merged into XFree86 4.3.0. - Conditionalized building xf86cfg, and disabled it by default, as we are moving to our new redhat-config-xfree86 tool for X configuration. - Backported apm driver from XFree86 CVS (4.2.99.1), DPMS support enhancements, and a few accel fixes. apm-driver-update-cvs-20020617 - Backported ark driver from XFree86 CVS (4.2.99.1) ark-driver-update-cvs-20020617 - Backported chips driver from XFree86 CVS (4.2.99.1), with hardware mouse cursor and 2D accleration fixes.- Added workaround for buggy laptop BIOSs (in particular the Dell Latitude C400) which use the Intel i830/i830m chipset, and limit the video memory to 1Mb of stolen system memory without making it configureable. This workaround addresses bugzilla bug (#65661) (Update: subsequently removed as it did not solve the problem.) - Removed currently unused Glide3Libver macro.- Fixed a bug in the VESA driver, where the X server would crash with an FPE due to a division by zero in miscreeninit() when the DisplaySize option was used in the config file. Inside xf86SetDpi(), pScrn->virtualX and pScrn->virtualY are used only when DisplaySize is given in the config file, however these struct members were not initialized properly in the vesa driver prior to xf86SetDpi() being called. Fixes bug (#66009) - Removed specfile BuildXwrapper conditional as XFree86 4.x does not need or use Xwrapper - it only existed for XFree86 3.3.6 in the past. - Removed specfile Build7x conditional as it was both poorly named, and the functionality it provided is no longer needed.- Added alpha-ia64-_LP64-fix to workaround issue in xc/lib/X11/XlcDL.c preventing X from building on alpha/ia64 with gcc 3.1. - Added 4 patches from Hewlett Packard for ia64 (XFree86-4.2.0-int10-hplso.diff, XFree86-4.2.0-pci-hplso.diff, XFree86-4.2.0-primpci-hplso.diff, XFree86-4.2.0-ps2-hplso.diff- Added explicit Requires: _x11bindir/xauth to XFree86 main package now that xauth is split out.- Update BuildPreReq: freetype-devel >= 2.0.6-3- Disable debug mode in cirrus alpine driver - filling logfile (#65704)- Split out xauth into a separate subpackage so things like ssh do not require the entire dependancy chain of XFree86 be installed merely to use ssh or similar (#38409,62426) - Added neomagic-Xv-support (#63609) - Added fix for deadkey-quotedbl in ISO8859-15 (#50282)- Updated parallelmake-libfont Imake patch for more parallelism - Fixed nasty problem in post script for libGLU, and added postun scripts that call ldconfig for the 2 Mesa subpackages - Added remove-bitmap-scaler patch disabled by default pending testing, compatibility, etc.- Added rpm post scripts for XFree86-Mesa-libGL, XFree86-Mesa-libGLU that call ldconfig properly- Fixed defattr for new XFree86-Mesa-libGL* packages - Removed Requires: XFree86 dep from Mesa-libGL subpackage- Updated parallel make patch from H.J. Lu!!! - Split out libGL.so.* and libGLU.so.* into new subpackages XFree86-Mesa-libGL and XFree86-Mesa-libGLU as suggested by H.J. Lu (#65152) and other vendors - Moved _x11dir et al defines to top of spec file to ensure their definition precedes their usage. - Completely removed XFree86.spec support for building using an external Mesa instead of the one included with XFree86.- Changed stripping policy and fixed a redundancy in the spec stripping- Bumped release number to -50 and rebuilt in new build environment- Changed Xnest, Xvfb to be in group "User Interface/X" (#62820) - Added fix for glint driver for bug (#60895)- Fix for i810 Xv maximum video size (#53329) - Changed += to =+ on various entries in our default XftConfig as per (#64970)- Changed all chkfontpath calls in specfile post scripts to use :unscaled for all bitmap fonts, otherwise scaled bitmap fonts get used, and that really really sucks. Makes Windows users run fast back to Billy. - Disabled Radeon PCI DRI code, as it has been reported as not working by a user despite claims before by others that it works. - Fixed euro support for pl keymaps as per bug (#64559) - Added Requires: /lib/cpp to XFree86 package as xrdb requires /lib/cpp at runtime and will fail with an error if it is not present. - Added fix for Rage 128 DRI related bug ati-r128-indirectBuffer from Michal Daenzer.- Removed s390 patch 3000, and fixed up file lists for s390/s390x- Replaced /usr/lib throughout specfile with _libdir for compat with architectures which have /usr/lib and /usr/lib64. - Replaced /usr/X11R6 throughout specfile with _x11dir - part of a gradual move to implement RPMified variable ProjectRoot support for developmental parallel installation. - Added the following patches backported from XFree86 CVS for AMD x86_64 AMD-x86_64-elfloader, AMD-x86_64-imake, AMD-x86_64-compiler_h, AMD-x86_64-int10, AMD-x86_64-misc-small-fixes, AMD-x86_64-os-support- Added AMD x86-64 architecture (Hammer/Opteron) support to the Imake config files.- Changed BuildRequires to pam-devel instead of pam header file. - Reworked BuildXF86DRI portion of host.def to exclude DRI on PPC for now - Reworked Glide3 BuildRequires so that the CURRENT Glide3 packages are required instead of the mess of conditional code that was there. People who rebuild now require the rawhide Glide3. Hopefully they read my changelog messages. - Changed xfs postinstall script to conditionalize the login shell at runtime instead of at build time to help ensure that packages _built_ on a newer release will work on an older release. Also, /bin/nologin which was added in January - does not exist, it is /sbin/nologin. Fixed, and tested. - Added utempter to BuildRequires and Requires - BuildXterm3d macro causes a buildrequires loop, and needs to be set to "0" if bootstrapping XFree86 on a new architecture. Comment added.- Fixed bug in XFree86 post install script, renaming XF86Config-4 config file to .rpmsave, and preserving permissions of the file. (#63822) - Corrected perms on /dev/dri- Updated mitshm patch- Added code to XFree86 post install script to remove xie and pex5 module load lines from XF86Config-4 upon upgrade/install since XIE and PEX are no longer provided nor supported, and X will not start if those lines exist in the config file. (#63119) - Added fix for sessreg being broken with high UIDs due to open() being used to open a file larger than 32bit offsets can handle. (#42850,#63116) - Fixed startx ugliness issue with hostname command (#61278,63117) - Removed double ghost flag in truetype-fonts package - Removed config(noreplace) from Speedo fonts.dir (#63357,63735) - I went outside today, and the snow is almost completely gone!- Made specfile remove fonts.dir files generated by XFree86 build, and "touch" them instead so they satisfy the ghost flag. Also for the fonts.alias for the TTF directory. - Removed CID directory from list of dirs to run mkfontdir in and add to fontpath, because our XFree86 build is not including fonts in there anymore, so it should not add the dir to the fontpath, nor run mkfontdir in it. - Fixed bug in truetype-fonts subpackage which was adding cyrillic to fontpath instead of the TTF directory due to cut and paste error. - Audited all font subpackage file lists for accuracy, correcting various ghost/config/verify flags, adding missing ghosted fonts.alias, etc. files to the lists also, so the files are owned by the package if a user creates them. - Added ISO8859-15 locale aliases to locale.alias for en_US and en_GB (#62844)- Added patch to fix various ATI Radeon issues. Will fill in the details of what specifically is fixed after testing the fixes. VTswitch is one... - Changed xf86cfg dep on Xconfigurator to /usr/X11R6/lib/X11/Cards- Added fix for Trident 9397 bug (#62119)- Updated ttmkfdir-freetype2-port-plus-cjk-chocolaty-goodness.patch to build properly with gcc 3.x (Bero,Yu Shao)- Fixed Matrox G550 cursor on second head mga-g550-cursor patch - Created new support for Euro currency symbol for gb, hu, hu_US, pl, pl2, tr, us_intl keymaps based on http://www.microsoft.com/typography/faq/faq12.htm, also closing bugs (#56818, and a few others I need to track down in bugzilla) - Put fonts.alias back in misc fonts directory, it slipped out somehow.- Flagged encodings.dir, fonts.dir, fonts.scale files properly for the misc, Type1, and Speedo directories in base-fonts, and changed the file globs to match just the font files in order to fix bug (#59619) - Fixed similar issues to above in bug (#61940) - Added XFree86-4.2.0-sparc_pci_domains.patch for sparc support (#62037) - Fixed duplicate Slovak entries - MF bug (#61709) - Confirmed ulT1mo fonts work and closed - MF bug (#61694) and (#62036)- Added ko, zh_CN, zh_TW entries to XftConfig for bug (#61282) - Changed antialiasing on small fonts from "any size < 14" to "any size < 11" - Added gb18030-enc patch from Yu Shao - ttmkfdir ported to freetype2, plus misc enhancements for CJK- Moved GL include files into /usr/include/GL for compliance with defacto and possibly official standards due to user complaints. - Changed host.def to not symlink GL include dir into /usr/include- Updated package descriptions to not suck quite so bad, and not be so meaninglessly verbose as they have been for a while. We do not want to scare away Harry Homeowner.- Cleaned up xfs initscript to use simpler syntax, and fix a few issues- Added the /usr/X11R6/lib/X11/fonts/TTF dir to XftConfig (#60901) - Updated bugfix for (#60906) from Keith Packard - Built on all arches- Fixed up tdfx DRI disable patch for hires on voodoo3/banshee - Updated Glide3 dependancy, to pick up new Glide3 package with libs directly in /usr/lib, instead of in /usr/lib/glide3- Added XFree86-4.2.0-mitshm.patch to fix security issue - Added fix for time wraparound bug (#60906)- Resolved some Brazillian and Icelandic keyboard issues (hopefully).- Added XFree86-4.2.0-xkb-is.patch to fix bug (#60594) - Moved all xkb symbol patches to section 800 - 819 - disabled tdfx font corruption patch- Added XFree86-4.2.0-ati-mach64-rage-XL.patch to fix problems reported on many Dell, IBM servers (#58188) - Fixed memleak in Xlib with XFree86-4.2.0-xlib-memleak.patch- Reorganized various fonts.dir, fonts.alias, encodings.dir files to try and fix some of the reported problems that have come up after various installations. All files are believed to be properly flaged with ghost, verify and config attributes now.- Fixed xfs initscript to use "reload" instead of "restart" for the condrestart option.- Fixed ghost/config flags on various fonts.dir, fonts.alias, encodings.dir files in file list section. - Fixed bug in truetype-fonts scripts which contained %_ttmkfdir instead of just plain old ttmkfdir- Added specfile URL to http://www.xfree86.org - Rebuilt with new built toolchain in new environment- Enabled Radeon PCI pcigart support in radeon_dri.c and radeon_cp.c - Disabled the ugly black and white pattern that X starts up with by default with XFree86-4.2.0-die-ugly-pattern-die-die-die.patch - blue or something else will replace it in the future likely.- Ghosted encodings.dir files in 100dpi and 75dpi font subpackages - Modified XftConfig with some suggested changes from Keith Packard for better AA font support, by moving stuff around (#59192) - Added Luxi font aliases, etc. to XftConfig for bug (#59708)- Created support for ATI Radeon Mobility M7 (LX) and also updated the pcitable, and Cards databases, in the hwdata package also. - Removed the conditional BuildPrereq for pam-devel et al. and replaced it with /usr/include/security/pam_modules.h which will automatically do the right thing regardless of what distro release is being built for.- Drop XFree86-4.0.2-neomagic_swcursor.patch previously included for sony laptops using Neomagic 256XL+ (NM2380), and set the option to default to on in the Cards database instead. - Created sis-swcursor option patch - Created xkb-us_intl-missing-commas to fix bug (#59529)- Experimental patch to disable DRI on tdfx when an incompatible hires video mode is selected that is known to cause unstable behaviour. - Removed -e arguments from calls to mkfontdir in spec file as they appear to be unnecessary now. - Added XFree86-4.2.0-zh_TW-xcin.patch to fix xcin (Owen Taylor, Leon Ho)- First merge with upstream stable branch xf-4_2-branch-patch-2002-02-06 - Misc fixes: Fix startx script for ksh, temporary workaround for IA-64, fix SIGSEGV when printing modes that have no flags, fix an unresolved symbol in libGLU.so with gcc 3, revert the composite sync default to off in the ati driver.- Updated XFree86-4.2.0-gb18030-20020207.patch from Yu Shao - Added i810-dont-force-XvMC-on-DRI-users patch to fix bug in where i810 users must enable XvMC in addition to DRI in order for DRI to work. (#59318) - Added patch XFree86-4.2.0-FIRSTINDEX-breaks-ttmkfdir to remove FIRSTINDEX entries from large font encoding files as it makes ttmkfdir deep six.- Added XFree86-4.2.0-i810-dont-use-empty-for-loops-for-delays.patch which removes null for loops from the i810 driver and i810 and i830 kernel DRM drivers for Linux and BSD. The for loops were used as a timing delay, and are now replaced with "usleep(1);" instead. - Added ttmkfdir-CJK-bug-54087-fix and bugzilla-54087-fix-gbk-encoding patches to fix bug (#54087) - Fixed libGLU symlink screwup by changing from absolute to relative symlinks- NEW XFree86-truetype-fonts subpackage which includes the new Luxi fonts (formerly Lucidux) by Bigelow and Holmes. - Added /usr/lib/libGLU.so.* and /usr/lib/libGLU.so symlinks to the libGLU libraries included in XFree86 in the /usr/X11R6 directories for OpenGL standards compliance, and disabled the non-working Imake patch I made earlier to do the same. This is fairly CRITICAL as any OpenGL software which uses GLU, will fail to compile without the proper links. - Removed a bunch of leftover Glide3 specfile mess. - Removed ancient no longer necessary hardlink between xterm and nxterm- Created support for interlaced video modes on 3dfx Voodoo 3/4/5, with the enable-interlaced-modes patch. - Created tdfx-fix-vtswitch-font-corruption patch to try to fix the VGA font corruption that occurs on VT switch on Voodoo 3 and other 3dfx cards.- Added XFree86-4.2.0-tdfx-should-be-2048-not-2046.patch to fix a typo I found in the tdfx driver. - Fixed bug in ttmkfdir Makefile. In RHL 7.1 & 7.2 freetype 1 headers were in /usr/include/freetype however in RHL 7.2 for some god known reason they moved to /usr/include/freetype1/freetype. No problem in RHL 7.2 as ttmkfdir used the freetype headers as part of the freetype build process rather than the system headers.- Fixed base-fonts and font-utils conflicts lines - Bumped release to 4.2.0-6, and rebuilt - RHL 7.2 build release tag 4.2.0-0.6.13- NEW XFree86-font-utils subpackage created for mkfontdir and ttmkfdir to reside in. This will allow these utils to be available without requiring other XFree86 libs or packages being installed. - Added patch to fix Matrox driver when VT switching to non-standard text modes (#59119) - Created new Red Hat custom modelines patch, merged previous patches into it, and added new modelines from bug (#53364) - RHL 7.2 build release tag 4.2.0-0.6.11- Created XFree86-4.2.0-xfs-syslog.patch to make syslog log to LOG_DAEMON instead of LOG_LOCAL0 - which is supposed to be reserved for local system usage only. Fixes bug (#24873) - RHL 7.2 build release tag 4.2.0-0.6.10- Enabled the now compiling gb18030 support patch - RHL 7.2 build release tag 4.2.0-0.6.9- Added gb18030 support patch to 4.2.0 , but disabled it as it does not properly compile yet. - enabled patch to force xfs to chdir to / before fork() - RHL 7.2 build release tag 4.2.0-0.6.8- Added XFree86-4.2.0-xkb-ar.patch for Arabic support (#56897) - Also built 4.2.0-0.6.7 for RHL 7.2- Added missing defattr to files list for XFree86-base-fonts (#59014,59020) - Moved all /usr/X11R6/lib/modules files/dirs into same section of specfile and added dir flags for dirs that were not owned by the package. - Enhanced the xfs initscript so "restart" shuts down xfs and starts it back up again, using the stop/start functions. "reload" is now what reload and restart used to be - "kill -USR1" of xfs. - Added zh_CN-i18n patch to fix an i18n problem with Chinese locales- Added chmod kludge to bottom of install section to force locale/common *.so.* libraries to have execute perms. - Also built 4.2.0-0.6.5 for RHL 7.2 - Removed the version from the Glide/Glide-devel build dependancies, so X will build if any version of Glide3 is present.- Added link-glu-to-usrinclude patch for LSB conformance. - Made xfs post script use /bin/false in 7.x builds and /bin/nologin in rawhide builds. (#53088)- NEW XFree86-base-fonts subpackage created to hold all of the base fonts that were previously contained in the main XFree86 package. The new package contains Conflicts lines to minimize upgrading difficulties. - Added Prereq: /sbin/chkconfig for XFree86-xfs - Removed XFree86-4.1.0-ati-r128-PR-fix.patch as it was no longer needed, although not harmful in either case. - Removed various other spurious patches that were no longer being applied, but were left in the specfile. - Removed conditional support for building XIE and PEX as they have been deprecated by XFree86.org for quite some time now and are now obsolete- Toggled BuildGlide3 flag so that Glide3 does not get built inside XFree86 specfile by default. Glide3 has been separated back out to a separate RPM package to simplify the XFree86 spec file, and ease maintenance.- Fixed bug in xfs initscript fixing a problem with xfs having a CWD under /usr, which when located on an NFS mount, causes problems. (#57110) - Enabled libGLw in build (#41791,#41974) - Built also 4.2.0-0.6.1 for RHL 7.2- Updated Wacom USB tablet driver Wacom-USB-driver-a25-update (#52947) - Added xkb-french-canadian keyboard to xkb list (#52309) - Fixed Xterm resource bugs (#48783,#49315,#58031,#58713)- Disabled the old DisableModuleStringMerging hack in specfile as 4.2.0 detects wether or not this option is required or not now via Imake tests. - Conditionalized pam and automake buildprereqs- Fix build in current tree (Glide is a mess)- Bumped release number to -2.0 with no changes from -0.3 package, for building into rawhide. All packages built for RHL 7.x should use release numbers less than -2, and preferably less than -1, ie: -0.x.y, so that upgrades stand a better chance of working in the future.- Updated the "Requires: kernel-drm = 4.2.0" to ensure the user has the proper DRM version installed. Newer rawhide kernels will now provide both kernel-drm = 4.2.0, as well as kernel-drm = 4.1.0 since the new DRM is supposed to be backward compatible.- Added missing defattr(-,root,root) to xtrap-clients subpackage- XFree86 4.2.0 released - first build- CVS trunk update to 4.1.99.7 - dropped patches that have been merged upstream- CVS trunk update- CVS trunk update- Integrated PPC arch changes from streeter@redhat.com - Integrated the Chinese gb18030 support patches received from yshao@redhat.com - Added SiS DRM patch (doesnt affect build as our DRM is in kernel rpm)- Removed eurofonts-X11.tar.bz2, and ulT1mo-beta-1.0.tar.gz from XFree86 package, moving them to their own separate packages that do not conflict with XFree86 namespaces. (Removed the ISO8859-2-Type1-fonts package) - Added new shared libs libXTrap, libXrandr.so to file lists - Added new applications to filelist: xrandr, xedit, luit, and various others, and their documentation. - Added new subpackage XFree86-xtrap-clients to hold the new xtrap client utilities- Made XFree86 server binary SUID root when not supporting Xwrapper (#56986)- Conditionalized out Xwrapper as its functionality is built directly into the X server now. To build it set BuildXwrapper to 1 in specfile.- Fixed that annoying Radeon hardware mouse cursor bug, where any mouse pointer that has a hotspot *NOT* on the left side of the pointer, caused the mouse to freeze up and jump around on left side of screen. XFree86-4.1.0-radeon-hardware-cursor-fix.patch- Updated xman patch, to use FreeBSD style man.config (why? I dunno.)- Added new entries to XftConfig for "console" and "fixed" to not antialias, which solves problems reported in konsole and other Qt applications. - Updated CVS source to current. - Removed patches that are now integrated into new upstream source base: XFree86-4.1.0-ati-radeon-dri-buildfix.patch- Fixed problems with WithExternalMesa disabled, in the file lists, etc. - Made new section for internationalizaton patches (30-39), and moved all locale.alias patches to that section. - Created some new aliases for en_AU, and en_CA and sent patches upstream.- Initial fork from 4.1.0 codebase to separate branch for 4.1.99.2 CVS - Set Build_7x to 0, which now means XIE, PEX and external Mesa are gone, and the doctor says I can stop taking my medication now. - Added preliminary support for building CVS snapshots, which will eventually result in nightly builds from CVS or somesuch. - Removed patches that are integrated into new upstream source base: ati-r128-SM, cirrus-foobaroo-fix, allow-module-flags, Xrender-required-libs, libXfont-put-eof, twm-coredumpfix, compose-iso8859-2, crash-message-fixup, xinerama-extmod-fix - Updated locale-nn_NO locale-en_BG patch, and sent upstream _again_. - Updated mkfontdir-perms - Added patch to remove SuperProbe leftovers.- Added XFree86-4.1.0-cpp-s390x.patch which fixes a man-pages of size 0 bug on s390 (#54927,#54932)- Disabled VT switch patch from -4 release after feedback of regression - Updated the Savage driver with Tim Roberts 1.1.20t driver, this should fix the reported bugs with Staroffice locking up X, and also video memory detection problems: (#31879) - Added XFree86-4.1.0-keycodes-abnt2.patch to fix abnt2 support bug(#55128)- Added patch to fix VT switch hangs that are hardware independent. This should fix numerous reported VT switch bug reports. - Fixed up host.def parallel make stuff.. HasParallelMake define is broken, and unneeded for now.- Removed __rmfontpath macro, and its usage, as well as the rpm 4.0.3 dep, because a lot of people were getting snagged on it, and it is questionable of the benefit of removing the fontpaths anyways, since it does not harm anything by leaving the font paths in config files really. - Changed /lib/libpam reqs to pam-devel instead - Commented the DisableModuleStringMerging specfile define, and moved to top of specfile where it is much more visible, as a lot of people have been getting snagged on this.- Modified and unified the "make World" buildlines by removing all the buildflags, etc. from them, and moving them into proper Imake configuration variables. - Added "-pipe" to default C compiler flags, as it should work on all architectures now. - Various other specfile cleanups. - Created support for ATI Rage 128 PP (1002:5050) XFree86-4.1.0-ati-r128-PP.patch- Silenced some Glide3 compiler warnings with Glide3-redhat-cleanup-1.patch.- Primary integration of changes in specfile to allow simple conditional build of fully debuggable modularized XFree86 packages. This is done by enabling DebuggableBuild RPM macro at top of specfile, adding "dbg" to Release tag, and rebuilding. The resulting RPM packages total in at 127Mb of data, as opposed to 55Mb or so regularly. In order to debug this debuggable XFree86, a special patched version of gdb 5.0 is required, which is currently a work in progress based on a much older patch for gdb 4.18.- Changed xfs initscript to use find -cnewer instead of -newer to fix bug (#53737)- Reordered font paths in xfs catalogue to put 75dpi first, etc.. - fixed couple critical specfile bugs that have slipped through til now- Fixed fs/config bug (MF#53103) as well as doing the same for XftConfig- Fixed a couple SHOULD-FIX bugs - fontpath order (SF#52413), metaSendsEscape bug (SF#49315)- Added "Requires: kernel-drm = 4.1.0" so that a matching kernel RPM package is installed and DRI works properly. (Brought up by mkj/arjan, and agreed upon by me)- Added dont-build-pex-docs patch from Branden Robinson to disable building PEX documentation (deprecated and obsolete) - Fix permissions on XftConfig, and other source files. (MF#52735) - Bravely bumping version to 4.1.0-1 - Added xinerama-extmod patch for bug (#52842)- Added Obsoletes: XFree86-ISO8859-2-Type1-fonts < 4.1.0, and provides ISO8859-2-Type1-fonts, to fix upgrade from 7.1, and prepare for future packaging. - Removed (noreplace) from fs/config, I *want* it to replace.- Completed font related fixups (hopefully) - Added specfile buildroot install time removal of fonts.dir files as they should never ever be included in the RPM packaging, but generated at runtime instead.- Added XFree86-4.1.0-ati-pciids-cleanup.patch, which includes misc cleanups to the PCI tables for ATI hardware. - Added ati-radeon-dri-buildfix from notting@redhat.com which hopefully fixes bug (#51884)- Completed font screwup overhaul take1.- Added XFree86-4.1.0-s3virge-mx.patch to fix some corruption issues on S3 Virge MX hardware. - BuildPreReq: rpm >= 4.0.3 (parameterized macro fix)- Exclude XFree86 doc dir from, and add xdm pixmaps to s390 arch.- Added xman-soelim-fix to fix bug (#51037). - Modified cirrus-foobaroo patch to disable debug messages (#51179) - Backed out an xfs change that breaks serving 3.3.6 servers. Attempted workaround for (#) - Disable ATI update patch to see if it fixes Radeon DRI bugs.- Added XFree86-4.1.0-radeon-slowdownthere.patch to fix bug where ATI Radeon hangs on faster computers when VT switching. - Created fix for mkfontdir to force permissions on newly created fonts.dir and encodings.dir files to be mode 644, fixes bug (#51636) - Added fix for iso8859-2 compose file (affects Mozilla) - Added patch to prevent X error messages from referencing log file if it has not yet been opened. - Created support for ATI Rage 128 SM (Xpert2000), and added rage128-SM patch to fix (#34732,#51230,#50231) - Created XFree86-4.1.0-locale-alias-en_GB.patch to fix (#47247,51169) - Moved video hardware documentation in /usr/X11R6/lib/X11/doc to main XFree86 package as that is where the drivers are (#50828) - Updated XFree86 dependancy on Mesa to -> 3:3.4.2-5- Synced with XFree86 stable branch to pick up upstream bugfixes - Added support for Japanese 109 key USB keyboard- merged new changes into specfile - added nn_NO, en_GB locale.alias support - Added support to xfs.init for .ttc font files, and to use case insensitive regex when grepping for truetype fonts. (#41169)- port & apply necessary DRI fixes that fell out- Added custom Red Hat XftConfig to fix AA font problems (#49937) - Added RPM_BUILD_ROOT to ia64 libxrx removal.. ;o(- Enabled "HasLinuxInput YES" to force wacom driver to get built with USB support with 2.4 kernel. Fixes bug (#50453).- Attempted fix for (#50043) -frame option ignored when -id option used on xwd (XFree86-4.1.0-xwd-frame-option.patch) - Made XF86Config-4 outside ifnarch s390 as Bero claims s390 uses it.- Added improved X startup messages for debugging purposes - Updated patch148 from ATI plus 2 minor buildfixes- Added XFree86-4.1.0-libXfont-put-eof.patch to fix bug (#48830) where the 100dpi/75dpi helvBO12.pcf.gz fonts were getting munged. - Added XFree86-4.1.0-twm-coredumpfix.patch to fix (#49674) - Added XFree86-4.1.0-cirrus-foobaroo-fix.patch from Egbert Eich to attempt to fix the Cirrus Logic problems. - Fix libxrx problem on ia64 (#49990) by removing libxrx from non-x86/alpha- Added XF86Config and XF86Config-4 as owned config files (latter ghosted) - Fixed s390 Buildprereq on gcc by unnegating macro test - Apply patch 3003 (64bit.patch) to all 64bit arch's.- Worked around rpm parameterized macro processing bug that caused a problem with postun scripts - discovered by rpmlint. - Reformatted all pre/post/postun/etc. scripts to be more readable and maintainable. - Added patch8 to xterm resources to provide RFE (#48783)- Added bison, zlib-devel as Buildprereqs - discovered on s390 by - Added ati-radeon-ve patch from ATI (patch148) - Made Prereq's on /usr/sbin/chkfontpath instead of on chkfontpath- ifnarchd patch1013 for s390- Created new macro __rmfontpath to be called from font package postun scripts to fix bug (#49243)- Added /usr/X11R6/man/man6, and other non existing manX dirs to XFree86-libs package so anything using them ends up having the dir owned by something. - Removed unneeded shared lib stuff (Xv/Xxf86dga)- Added BuildPrereq: gcc >= 2.96-93 to guarantee compiler used supports the -fno-merge-constants option - Made External Mesa usage conditionalized via WithExternalMesa so that it is easy to rebuild XFree86 with/without using external Mesa, which will be the default at some point in the future.- Explicitly state mkfontdir path in __mkfontdir macro (#48829)- Removed SharedLibXdmGreet from host.def, it is the same as default anyway - Removed tcl and tk as buildprereq's, Bero says they're not needed anymore- don't require XFree86-compat-libs in XFree86-libs- Fixed missing requiredlibs on shared libXrender so that prelinking works. - Built 4.1.0-0.8.4 - Woops, I left shared libXv and libXxf86dga on again by accident. - Built 4.1.0-0.8.5- Fixed up xdm pre script - Fixed up missing Latin2 Type1 fonts, they are now in the fonts/Type1 dir - Removed shared libXv.so and libXxf86dga.so at request of XFree86 team, and in conjunction with discussion with Debian X maintainer.- Fixed ia64 Glide3 build failure.- Merged with xf-4_1-branch to pick up stable bugfixes, and removed XFree86-4.1.0-redhat-glint-missing-symbols.patch, also Banshee ramtiming fix picked up from the update.- Really disabled string merging this time - Added verify flag to /etc/X11/fs/config - Disabled shared libraries Xv/xf86dga for standards compliance. These libraries do not have a stable API and should not be shared. It was a Bad Idea(tm) to ship these.- Disabled string merging in modules using ModuleLdFlags in host.def - Added WORLDOPTS= to make World to get rid of the "-k" passed to make so that failed X builds will fail immediately instead of continuing on.- Phase 1 - Amalgamated all my local font related changes to get fonts working better in the new 4.1.0 release. Straightened out 75dpi/100dpi brokenness and split up packages. - mkfontdir is now called in post(un)install scripts, and fonts.dir files are not installed as part of packages anymore because it was a bad broken idea to do so because these files should be generated, not provided. - Added missing \ in main package post script- Added customer requested custom 1152x864x85Hz modeline- Reworked all ISO8859-* font subpackages for 4.1.0. fonts.dir is not included in the packages, rather it is generated at post install time. - Refixed (#31618)- More s390 cleanups- Integrated s390 patches sent by Than from Oliver Paukstadt- Cleaned up unneeded Freetype2 stuff in host.def- Take-1 at cleaning up some of the font related 'issues' due to the different way fonts are generated in 4.1.0. Still many issues remaining.- Folding the external Mesa into XFree86 created as many problems as it solved. The DRI CVS trunk has a more permanent solution worth checking out. - Since the V4L drivers have apparently been included in the main XFree86 package all this time, I've removed the separate package.- First stab at assimilating Mesa 3.4.2 package into XFree86 package - Added missing symbols to glint driver (#42751)- Rebuilt against Mesa 3.4.2 in new environment- XFree86 4.1.0 released - built 4.1.0-0.0.1- Remove GLw manpages since we don't ship libGLw (#41791) - I finally got XIE and PEX disabled completely for sure this time absolutely no question. No really... It's gone.- Updated to 4.0.99.901, removed unneeded patches - Removed Glide3 Provides/Obsoletes from devel package, and removed the Glide3 devel stuff from that package also - only X needs it during build - Removed BuildGlide3 conditional - Glide3 will always be built from now on - Added conditionals BuildPEX & BuildXIE. These have officially been obsoleted by XFree86, and are no longer actively supported or used by anything.- Added "Options" database file, and also the following apps and their manpages: dpsexec, dpsinfo, texteroids, xftcache - Removed hardcopy doc creation code (groff) from specfile. The XFree86 makefiles have been fixed now to do this themselves properly. - Removed Glide3 headers from XFree86-devel. Only needed during build, no apps should be linking to Glide3. - Simplified filelist for libs subpackage by using globs instead of hard coded library version numbers. - Removed use-cpp patch, done with imake now.- Merged bero's changes to my local sources, and changed libXmuu inclusion from libXmuu.* to libXmuu.1.0, so that the *.so doesn't get picked up also, as it should be in the devel package.- Add missing libXmuu.* files- Many fixes from H. J. Lu including enhanced parallel make, xterm-xaw3d, pex patches, specfile updates, build enhancements and a partridge in a pear tree. ;o) - Fixed insane off-the-screen indenting levels in xfs.init- Updated to 4.0.99.900 - Updated with the latest parallelmake patch from H. J. Lu - Added new specfile option ParallelBuild, so that any parallel build stuff can be conditionally defined in the specfile in case the parallel make patch breaks in the future.- Fix build- Updated 4.0.99.x package to 4.0.99.3 release, forward ported all relevant changes that have occured in 4.0.3 packaging since March 27th until now, and brought the 4.0.3 specfile changelogs into the 4.0.99.3 spec changelog. The two releases should be syncronized now.- Syncronize XFree86 with Mesa 3.4.1-1 - Disable alpha-hack from May 3rd and add new patch from Compaq to replace it- Updated patch0 to latest XFree86 stable branch for 4.0.3. Fixes many 3dfx bugs in tdfx driver. - Moved locale data from /usr/X11R6/lib/X11/locale in main package to XFree86-libs package where it belongs. (#38408) - Corrected dumb reversal in XFree86CustomVersion Imake define in specfile- Reverted ia64-int10 patch from Alpha. - Added XFree86-4.0.3-alpha-hack-bypass-int10bug.patch from Jay Estabrook for Alpha.- Removed RPM_BUILD_ROOT from glide section (#38031) - Enabled XFree86-4.0.2-ia64-int10.patch on alpha as well for testing.- Added XFree86-4.0.3-compose-kill-ctrl-shift-t.patch to fix (#23473)- Added rhl-startupmessages patch to aide debugging and troubleshooting bug reports. This should increase the usefulness of user submitted XFree86 log files and decrease the turnaround time on requesting info and waiting for someone to provide that info. Patch doesn't do much yet, but will be enhanced later. - Added xset manpage enhancement- ifarch'd XFree86-4.0.2-agpgart-load.patch to only apply on non-alpha- Removed dependancy of main XFree86 package on xfs by moving the XFree86 server modules "/usr/X11R6/lib/modules/{codeconv,fonts}/*" from the xfs subpackage to the main XFree86 package as they never belonged in the xfs package in the first place. This is an ancient packaging error. - Added patch from Tim Waugh to fix Xaw password entry in vnc. (#19231)- Killed ancient non-FHS compliant symlink to /usr/man/X11 - Disabled XFree86-4.0.3-savage-glitch.patch to fix (#34376)- Made glidelink get called before ldconfig in %post libs- Simplified the main package %pre script symlink munge code - Fixed bug in %post xfs section - inadvertently added in 4.0.3-6 release - Disable patch #220 as it is unnecessary - taken care of via Imake ages ago.- Fixed (#32574) by adding a preinstall script on xdm which moves the dir /etc/X11/xdm/authdir to /var/lib/xdm/authdir if it exists, and symlinking /etc/X11/xdm/authdir to it. Corrected the filelist for xdm to include these changes. - Fixed alpha build failure with tga driver patch, ifdef with no macro.- Force enabled BuildXinerama and BuildXineramaLibrary in hosts.def - FHS compliant manpage sections (4x/7x)- Added new patches submitted by Compaq for various drivers, etc. XFree86-4.0.3-compaq-alpha.patch, XFree86-4.0.3-compaq-glint.patch XFree86-4.0.3-compaq-radeon.patch, XFree86-4.0.3-compaq-r128.patch XFree86-4.0.3-compaq-mga.patch, XFree86-4.0.3-compaq-tga.patch (courtesy of Jay Estabrook)- XFree86 4.0.3 defaults to building libfreetype if Xft is being built, so freetype is now disabled with Imake defines. - Removed shared libsXxf86dga and libXv hack I added on Mar 9 and did it properly with Imake defines. Also fixes (#31232)- Added patch XFree86-4.0.3-glide3-libdir.patch so tdfx driver can find the Glide3 libs included with X package. - Minor xfs.init fix - Added XFree86-4.0.3-mga-hwcursor-fix.patch- Updated g450 port from 4.0.3 to XFree86-4.0.99.1-mga-g450-dualhead.patch - Updated Savage driver dropin to version 1.15 - Added patches 405 - 408 from 4.0.3-4 build- Initial specfile fork of 4.0.99.1 from 4.0.3-3 - Removed unneeded patches: XFree86-4.0.1-dri-sym.patch, XFree86-4.0.2-mga-busmstr.patch, XFree86-4.0.2-r128-busmstr2.patch, XFree86-4.0.2-r128-mobility.patch, XFree86-4.0.2-agpgart-load.patch, XFree86-4.0.2-r128-drm-do-wait-for-fifo.patch - Updated patches: XFree86-4.0.99.1-parallelmake.patch, XFree86-4.0.99.1-locale-euro.patch, XFree86-4.0.99.1-winkey-alt-meta.patch- Drumroll...... Final (or is it?) working G450 dualhead update. ;o)- Added pam patch for xdm from Nalin Patch404- Fixed uninitialized memory reference in xlib (Patch401) (#32222) - Make libs subpackage depend on kudzu-devel, pciutils-devel for Glide3 (#32534) - Matrox G450 updates- Fixed glitch in %post xfs section of spec (#32281) - Set "umask 133" in xfs initscript so that fonts.dir/fonts.alias files created during script execution work also for non-root users that may have changed root's umask. - Updated Mesa dependancy in specfile to 3.4-13- Removed XFree86-4.0.1-xterm-appdefaults.patch as it was redundant- New XFree86 4.0.3 released and integrated. - Updated patches for 4.0.3, removed unneeded patches. - Built 4.0.3-1- Reverted G450 driver back to dwmw2's driver. - Built 4.0.2a-2- Added XFree86-4.0.2-locale.patch from Phil Knirsch which should fix all problems with euro locales, and other locale issues including bug id's (26300,21882,28447, others) - xfs crash errors should all be fixed now by the XFree86 stable branch patch. Bugzilla (#22538)- New patch XFree86-4.0.2a-mga-g450-dual_6.patch - G450 driver update from Luugi Marsan - Added new euro locale stuff from Phil Knirsch - Added XFree86-4.0.2-r128-drm-do-wait-for-fifo.patch from Gareth Hughes which should fix the r128 errors in (#26999)- Added provides entries for Glide3 in libs and devel subpackages - Take 2 of XFree86-4.0.2a-im-being-included-and-i-cant-get-up2.patch - Moved XftConfig file to /etc/X11 and made compatibility symlink.- Updated main tarball to 4.0.2a (even though it's the same code we had before with 4.0.2 + stable branch patch). - Updated stable branch patch to bring 4.0.2a up to date with xf-4_0_2a-branch - Nvidia nv driver updated with MANY fixes from XFree86 stable branch - Added shared libraries: libXv.so, libXxf86dga.so - Re-enabled Sparc build. That machine is slow, someone needs to put a spark under it to get it going. ;o) - Fixed (#13882) with XFree86-4.0.2-winkey-alt-meta.patch - Fixed (#19308) added X11R6-contrib obsolesence- fix glide dependencies- Fixed alpha/ia64 build problems - Built 4.0.2-12- Created XFree86-4.0.2-r128-mobility.patch to enable external display on Rage Mobility laptops, etc. - Added XFree86-4.0.2-xft-match.patch (from Bero) (#30836)- Added XFree86-4.0.2-alpha-pci-resource.patch (#30417)- Glide3 is back (was previously removed on Feb 22) - Removed loaderProcs.h hack, and replaced with a patch. - Tidied up a tonne of XFree build warnings - Added XFree86-4.0.2-im-being-included-and-i-cant-get-up.patch stop loaderProcs.h from being wrongly included in building Xnest due to a compiler bug.- Added mga and r128 busmaster patches that Todd Nix sent me from Kevin E Martin .- Updated xf-4_0_2-branch patch to todays latest stable code - Removed unneeded patches after update: XFree86-4.0.2-time.patch, XFree86-4.0.2-sis-unresolved-memcpy.patch - Added XFree86-4.0.2-agpgart-load.patch from Bill Nottingham to force load AGPgart kernel module when a DRI module is loaded. - Fixed "chmod" in xfs.init (#18573) - Added BuildPreReq ncurses-devel - Repaired ia64 int10 patch - Tidied up specfile, particularly the host.def generation section, and incorporated xfree86.def mess into host.def section where it belongs.- Added updated driver for Savage chipset. Should fix all known issues with Savage cards that I'm aware of. - Removed Prereq of chkfontpath on xfs subpackage and modified xfs initscript to only call chkfontpath if it is installed.- Added specfile %pre fix for (#26857) - problem upgrading to Wolverine from Red Hat 5.2 due to directory/symlink fu glitch. - Enabled Utempter for xterm with "UseUtempter" in specfile - Added neomagic softcursor patch for (Helge Deller ) - Built 4.0.2-10 - Created patch XFree86-4.0.2-sis-unresolved-memcpy.patch to fix bugs (#19530,24670)- Added spec hack to fix bad include of loaderProcs.h in miinitext.c I believe this fixes some of the r128 DRI problems. I can't get r128 DRI to crash anymore now anyways. It isn't mind blowingly fast though. May have fixed other problems as well. X seems much smoother now, and VT switching is much faster also.- Added ExcludeArch tag to quell failed sparc builds (unsupported) - Re-enabled DRI for mga/r128, but they still do not work- Updated fix for xfs death bug with truetype fonts. - Added chkfontpath require to xfs subpackage (#19603) - Added XFree86-4.0.2-banshee-ramtiming.patch to fix (#26351,#18810) - Added ia64-int10 patch. (Bill Nottingham)- Fixed xfs.config- Disabled DRI drivers r128 and mga because they are broken. - Built XFree-4.0.2-8 - Fixed bad noarch thing - Built XFree-4.0.2-9- Merged in separate XFree86-ISO8859-2 font package- Added 1400x1050 modeline patch for newer laptops - Misc fixups - Disabled DRM kernel module building as we don't use these modules anyway. - Updated 4.0.2-branch patch - Updated XFree86-4.0.2-xfree86.lst.patch- Internationalized xfs and xdm initscripts- New fix for Cirrus 5480 (H.J. Lu) - Matrox g450 patch that _hopefully_ works this time (from dwmw2)- (#25890,#26131) tdfx dri driver not included.. specfile mixup, fixed. - (#23896) Added option for building with Glide2 (default=off). Dunno if it works or not as I'm not testing it. If someone wants to further the enhancement at all, I'll accept specfile patches to add any missing bits. ;o) - (#25131) Fixed xfs.config file glitch (Truetypex typos), and a similar bug in specfile hack. - Updated to recent XFree86 stable bug fixes (Patch #1) - Fix temp files vulnerabilites in xman on systems with mkstemp() - Fix temp vulnerabilities in Xaw/MultiSrc.c - Fix temp file vulnerability in gccmakedep based on report from Alan Cox. - Fix temp file vulnerability in Imake.rules - Fix temp file vulnerability in InstallManPageAliases - Fix Neomagic 2200 screen corruption - Add an imake control for determining when xload should be installed set-gid- Added g450 patch from Matrox. Lets see some dualhead action baby! - Filtered out M$ Windows CR's from the Matrox patch, hope I got them all. Windows bad, Linux good. ;o) - Removed all prior g450 patches as the new one repaces them all - Many spec file simplifications, and general tidy up job.- fix perms on xload- chmod 644 XftConfig before piping to it- Disable Cirrus 5480 patch as it wont apply cleanly to 4.0.2. - Removed v4l patch - its in 4.0.2 already - (#23701) devel dependant on libs instead of main package. - Added TrueType font dirs to config files for the ttfonts package - Numerous other tweaks - XFree86-4.0.2-4 built- Fixed xfs initscript to only start at boot time if X is installed. - Put eurofonts package back and fixed it up with suggestions from H. J. Lu - as 2 font files are still needed. - Changed xload %install time chmod to %attr instead - Tidied up %install section, removing redundancies. - Added Cirrus Logic 5480 patch from from Egbert Eich (#24177)- Removed manpage "hack" section from specfile - Removed eurofont tarball entirely. It's included in X already under different names now. - Removed mga, ATI, mesa patches from yesterday as they fail the build requiring many other missing items.- Added ATI driver patch from XFree CVS - lots of new fixes - Added mga driver patch from XFree CVS (mainly for G450) - Added mesa patch from XFree CVS - Added BuildPrereq: pam-devel (Bill Crawford ) - Commented out BuildPrereq: pam-devel as it failed me - Remove eurofont .bdf files after pcf's are created to quell duplicate font errors.- Added Xft update from XFree CVS - Updated %clean section - Changed specfile bunzip references to bzcat - Fixed all specfile references of "tar x" to "tar xf -" - Added romanian-keyboard-fix from Cristian Gafton - ATI r128 update from ATI.- Disabled tdfx builds as they die with missing glide3 libs (at least in RHL 7.0 it does)- Added BuildPrereq: flex >= 2.5.4a - Added BuildPrereq: freetype-devel >= 2.0.1 - Added BuildPrereq: Glide3-devel >= 20001220 for tdfx- Updated fhs patch for 4.0.2 - Updated portuguese patch for 4.0.2- Integrated Matrox G450 driver for testing from Matrox- Added xfree86.lst patches from Brent Fox and also (#5977)- Misc spec file fixups- Integrated ATI driver from Michael Smith - Built 4.0.2 release 1- Add a couple of missing man pages, noted by Thayne Harbaugh - Use system freetype, now that we've updated - Add missing XftConfig file- don't list the driver modules explicitly; ship them all- integrate tdfx_dri into main XFree86 package- fix typos in xkb symlink/dir munging- Add missing Xft (Xrender) includes in -devel - Move xkb files back to their old locations, we can't replace directories with symlinks - Add /etc/X11/xkb -> /usr/X11R6/lib/X11/xkb symlink for compatibility - 4.0.2 - Get rid of twfix patch (it's in 4.0.2) - Patch 302 must be applied with -p0, not -p1 - Add a %clean section - Get rid of /etc/X11/xdm/xdm-config in XFree86-xdm. Since kdm uses it, as well, it should be in xinitrc- Added XFree86-4.0.1-twfix.patch from bhuang@redhat.com to fix zh_TW locale. - Added locale.alias patch from Jakub Jelinek. - Removed unneeded patches, and tidied spec file a bit (yeah right). ;o)- tweak the -pic patch so it works with the ELF loader (in theory) - don't try and build the kernel modules- reapply the parallelmake patch - hm, libX11 and libXmu are now .6.2- Updated to CVS version xf-4.0.1Z- add back in some patches (FHS, Xwrapper) - make it build on ia64 (hopefully)- Added patch XFree86-4.0.1h-xf86Wacom.c-kernelheaderclash.patch to fix 4.0.1h build failure.- Fixed bug (#21303) changed 2 uses of arch "i386" to "i386 i486 i586 i686 athlon" - Remove dependancy on tar 1.13.18 by calling bunzip2 and piping to tar - Removed I810Drm.tar.gz entirely as it smashes other dri code (mga,etc) and is ancient- Updated to CVS version xf-4.0.1h which includes many updates and integrates many Red Hat patches directly into XFree86 CVS.- Update to current. This should fix, among other things, DRI/DRM stuff on Matrox cards and i18n. - Fix build with tar 1.13.18 - Add XKB patch for Romanian keyboards from Christian Gafton - Fix up Euro sign support- Don't apply the PIC patch on non-ia64 platforms, it breaks X startup on x86.- minor tweaks for ia64/glibc-2.2 build- never mind, it doesn't. Wheee.- oops, the r128 mobility patch breaks regular r128 cards- fix DGA for real - add support for the r128 Mobility chips- fix xterm app-defaults file (#13584) - don't build a shared xdm greeter lib- fixes for MGA dri (text corruption) - spell Portuguese correctly in keymap list (#12343) - fix some absolute symlinks (#16133) - fix xman's man config path (#15974) - fix broken xdm pam code (#15952) - fix bug in ATI driver (#14344)- revert some of the i810 code; it breaks VC switching- fix DGA so that 4.x-compiled DGA1 apps can run against 3.3.x servers - fix tempfile handling in makedepend - use cpp, not gcc -E- more CVS updates (Neomagic fixes, some i810) - fix sparc build- do *not* do xfs stop/start in condrestat- update CVS (i810 fixes, others) - move some more files out of -xdm, into xinitrc package - remove /usr/share/fonts/default/TrueType from default xfs config (#14683) - fix weirdness in the i810 driver update- hack to work around gcc bug triggered by xwd (patch207) - modified xfs post script to add in no-listen = tcp per Bill's request- Some i810/i815 stuff (new I810Drm.tar.gz, updated to xf4.0.1-I810copy.patch only to find out that it is already included in this source tree).- add patch for xf86 vidmode extension - add icelandic keymap to list of keymaps- hey, we recognize the geforce2 now. Wonder if it works. :)- oops, I broke the sparc Mach64 patch on sparc- move initscript back - update CVS - remove Mesa (use its own separate package for easier updating) - kill the GL wrapper - fix linking of dri modules so that symbol resolution with Mesa works right - add sparc Mach64 patch from DaveM - fix sparc Mach64 patch so it builds - add ffb patch from Jakub - fix xfs user creation- Fix Mesa library version info - Set Mesa package version to reflect Mesa version (not the XFree86 version), and set release to something that will facilitate upgrades from older Mesa pkgs while still being informative. - Only pass -j to toplevel make, which takes care of passing it on down.- fixed xfs condrestart to do a real stop/start. - fonts.* files in font directories are config(noreplace)- Update Mesa to 3.2 release. - Use mmx/sse/x86 stuff for Mesa on 386 and up, not just if building for 686 (since MMX is available on 586 many times, and since the functionality is runtime-configured). - Try to integrate newer i810/i815 driver (unfinished)- update CVS while I'm here - don't use ix86 PCI code on ia64 - fix some int/long bogosity in some of the drivers- update CVS - build with new compiler to fix Xserver problems on i386 - add parallel make patch from H. J. Lu- oops, open the right software Mesa lib - build DRI everywhere- fix Mesa install - add Romanian keymap so keymaps get installed- add console.apps file for xserver (#13325) - prereq chkfontpath for font packages - update CVS - bump version to 4.0.1 - add lemming patch- update CVS, sparc patches now included - ship pam files - point X symlink to Xwrapper, not XFree86- fix pamsession, xfsredhat patches - turn off joystick, it relies on magic symbols defined *nowhere*- merged in a bunch of patches from 3.3.6 that somehow got dropped including: - xkb changes - some xfs changes - xterm app-defaults - pam session fix for xdm - define some paths in host.def instead of a wild mess of patches - kill compat links from /etc/X11/ to /usr/X11R6/lib/X11/- add Mesa wrapper - put Mesa libs in /usr/lib - return of Xwrapper - take setuid off XFree86- remove 3 obsolete SPARC patches - add 2 new ones- put setuid on XFree86 (temporary) - yank setuid off xload - put the xfs %pre back in - xfs initscript munging - fix ia64 patches- Update to current CVS- return of the font %post/%postuns - move some of the xdm stuff to xinitrc - put libglut in with Mesa- build glint module on alpha- Fix up Xaw (Bugs #11404 and #11426) - more FHS fixes (fix bug #11358 and some related unreported bugs)- split off Mesa and Mesa-devel - FHSify - fix build on alpha and ia64- merge in ia64 fixes- change PAM configuration to use system-auth - build fix for Sparc- don't define GlxStubLib if DllModules is not defined - otherwise things don't compile on i386- support ffb, leo and cg6 on sparc - only switch to dlopen() loader on selected architectures- switch to dlopen() loader- add man pages to -devel- more directory shuffles- more random fixes- various fixes - fix specfile for ia64 - merge drivers back into XFree86 package- update xf86cfg - fix up the xf86cfg patch - various fixes- Fix up x11perfcomp and move it to -tools - Fix up xfs (missing init script...) - chkconfig'ize xfs - move pswrap to -devel- Add xf86cfg (successor to XF86Setup, taken from the XFree86 4.0a tree with permission from the patch author) - Move the directory fixup stuff from %pre to %triggerpostun -- XFree86 < 3.9 We don't want to fix things twice...- Fix post script breakage - Disable build root strip policy (drivers can't be stripped) - strip the stuff we can safely strip manually - add /usr/include stuff - add some stuff from Mesa - we don't really want to obsolete Mesa without providing GLU and glut. - move some man pages from XFree86 to XFree86-tools, where they belong. - move some files to subpackages: beforelight, xclipboard, xclock, xfd, xieperf, xmag, xwininfo are now in XFree86-tools (rather than XFree86) makeg is now in XFree86-devel (rather than XFree86) - obsolete Mesa, glxMesa* and xpm in -libs - hope it still compiles- Even move sparc tweaks- Make it compile on sparc (Patch from Jakub Jelinek , some additional fixes from me) - Autodetect the right version of Tcl/Tk - remove xinitrc, we get it from the xinitrc package- More fixes - Fix compilation on alpha - TODO: Port our 3.3.6 patches- Add the man pages ;) - Merge some patches from the RPMs at seawood.org - Various fixes- Initial RPMmkxauthXFree86-xauthdolly.build.redhat.com 1125615106- |@?6.8.2-1.EL.13.16mkxauthxauthmkxauth.1x.gzxauth.1x.gz/usr/X11R6/bin//usr/X11R6/man/man1/-O2 -g -pipe -m32 -march=i386 -mtune=pentium4cpiogzip9i386i386-redhat-linux-gnuBourne shell script text executableELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), for GNU/Linux 2.2.5, dynamically linked (uses shared libs), strippeddirectoryRRRRRRRR? }{@T,,"TiC| WD-,E\$]ٲFYYYiK) D͊ ӔTdw~P{g̙3gΜ9g;$mVSbMZK֙ؤ8S<&b Y8v.s^F0lC5֔t~ڸSjM#5 m\4,@a&Z5vGO7dfAb)q侱Z._ eْkv-=Mgoi3@ݞmXB~qq+3VͰV2䱱7e0|Sgtִ صx-6ʚ+ yy0 B d18LCLa i碨`23ZmRf s ̙ 08t;@0f{ f@>+`l+2gBi}`tec4fgL `c)RV˒gIDYA1w`Uнg\eݽy@d4@3NÕfb5,` y&YYh7Gls4אdc |vG_\(B[ ;H=Ldb ρb#"HD @GHyl6j_lvlZu2ڬ*mZrrkY2l!˿ny\0d-NX]+SX̰i"#59Di5gd?N*A|e%ڕ,(T@v7xQ3Y1114Zcʯea(Uœ5Rђ] v; 2ł<ߒc-7`׎~G42uEUj?`$AZ&8ugnVL֭܈=?jun܀ /_BMv60eL,5CdbLz#`ivP >t'[o8f7:Ù u%nP?=4_n]gJ"OumG ^dlnn$JyBK m`'ӁAhi lo%3L RdF^5w蚌ț5ث%]n=X͠]YfZ#R+2O&-Y XT\b]`9!X un+L(y^GhO@oEB8C &qr"/o#2+ӄfMt& ך7pfRg2J;l4FFrujʢ kt }^ޫ>TE)&Z {a,33|` :0#&r)BRߕݸDFF^SFL4Ӡ &eG&ڪ=CEKȫӘפblYǒW5>͘ 3wBQ:?=@XRaADUBa\QR?~W7$I%.\asb2!`3F4̐ 8pC\ rBɈ.\l99h {0 E [d#4~_Q$$.EoGDF"ӟ~Oj#m-IՂ.|JzqNKjlgB`{;#;l?7<8ncx\+ۂ[ݣt#{F;Qo0~.!sf]IC0]u"|򎄵'[ʼ`i7权l9&fip` V_QD1sԢ5h"q>w, PdlWr&?m{wV(/-\Gr `ysy&V\ZɅ :X+Ou\S=fXVސoO?7Bzo2qT k2a/o_,ŸF@0˫#@WyjЀ94[Fb>W`񍇁C2ធ!&ljf8g!t|[ Ckmq.h}e#o6nK}B5iƧk׍:˜zy5},R?( 5\=rPV4pf%{}_*c[惉5uIƿk&|5zＲ7^G* "@q/(Up-)dDCڒ 'mh`=B  -T:B#/SpPaq.!X(hj t ],@Permx=HhAi_Ɵ\o^nV<,qbbY~(}Ęt.+0thދ6y~[X^~4F)t`Qc@L$C3n(V*ïp @8Kp  @OSJ1[oWԹέwk|\#;yN?^2u'Dyu=oG+:rcܶx/ _6/ `F_*UR*n `X5?bZsxzOw ӿrwtOӁ?q.]98`ze??CLyz/WqignY Ng!~R\^'c[_=Cy%^ޖ;yj߻1 <׏Y<@.N_  V^~1y~_s]xOwo)Z/YsHYEK}yz~a^~KlM~S+]UX>(gG0 ?L_|ͼ|Ko-(> 7 \˟7||u j|P~Uɼ\+7X';T܍0 #'И!i\k@&AO0 $$sC:^?!s$ޙ 7 O+8 N <\WpYWXP [Wp`ħ bBp,++ J +, uxe=^aހW0%x`a+^i.+&xz^!؏W* F^!`+8G _ ^a=WRQs4΀JƙSMO8#rjhigFq44Δ&˘.":cz^B+iz9L/t1WtMi4.Lut1=Ǵ7ǴKc:4R:nr:4??MOOӕtii??MWtii??MOO:4]OǟY:4Dǟ/i9M1#קH~*:0n9%B bΎ%J/(ŋG#=A )KI-8` u꺻gWZUxBuELlduVZp uՇ`fs<ePsՇ AF(eaJZ򜆞BaLIx.uT1QúHBf2 < T=m@V=}CF/}މ9?8q;&ǻ#*#Ȫ ؟ Q߰j8]wob/$Gh ')+JHR'6ms{sFUE\@gΎUnB]HhB;o MAiuQ;'0S] PeAB*[QXu+Bx~QEP8uF@h#W8Cx^ҫ* yl( 4 Dg=Yn^ ?xP/KKKDڎ&qh9 ¨Ir+,yR'QdZZjO\+4;mdy9;wAي&U@MDwl/eq5Q/)"')VSR*&`I*df6 wH*iRc8ҝ➌2`,$1w ' \Q=ǺaD+=-  K\@H(܄cby4Q\\#"57C⍆tm4Qz8%]449O7u3^wdahuhR\sUm*%r~'ܬ~K:ЍR=?S ~uXB_5T%B_EA{r齇) wX1Q{[#wJWSh!Tr5`Oݨ~}B.a;H =1saTbH#T҂)Q(5ǀ>T#b{h cOz5pxhן )ږ?E+,wd1R=Lx9?׈$J(Ŝhh4Džp:¹*tdkv6´2XU =U^0y- eemhiPwPGv|=L"mYosx=;ZԘQ Z<ؒgx瘫V!.i*eOwC־ˉ vvYP8@(s-Tj~Qéu= Z6=p67-bs7!*N2BB82L[ Vs(ΉRQ),?Y^HnجP>x?pj'7E}{<08CӮfL|# R`6o܋OORh_#VOAe$VL5yGLovE]&zJ u S7A.ؘO -WABEṖLJB*+( eX<>R,գxX3Y,#~y j8Tln#  -ŝ|*d))|q?E(`O9 ݊@!.d 8 P1<4##c>2 ntj1D 71էix:tYұhTރɻ}T9йFy0*Ɇl;8 /UhT!t o-љ"RG$D''Wͫg%+?KڎO`g;񫞜QiOmADz.`k>7Q^H)$-LBЇsc%n&A#j랒9%S56pUR[t*}%MuUImCRznrIȧX+mćIP%E~heÉ^ TtƼ;i}n69Ja?`H Pq)VIey/Vb䚮,e1bpR*>vhzS cݼ!PEiuxSh{B>LUL*/"~C 15/b2ֈPD7ʉ{<|-qxM`zzD3 mEA<@Z4?~ŷ|IoߌbPߒAwnon+SH CB[Pj<2j390v,^J)ڢ.q5=o(VEVxK"c>6 :w~\'_B9Cl.hB/Ǽ˔Qn\48/`\>ZG9 ]P^5hRF>2w9^ot͸"KÊ)qäO p <˨+X{:ȄcX@kX]°eOfB<'M?2Bˏs3Vf1lqH KWy*x_ZU9RorHix%YxIVa߫HCBDgMJ#o(\Ǎdݛ׼e`Пd#Oԙ(v e"rAR}(9=(X8V;inEB~Tr6HTe]hnshb/#?sI`LvMMB_7qf]J+oCz$IJiRj#>ۊ7S)5|Tc(\'޺'bzji4= vX$Ajpv>1}}٪ ;d!(JQR]U $.7RvE,=>, juQ'/_52MD}׈|6c(^⒤䏨`mpGLj@A0%Bi 8# XFJ#R~Oe:qRZ+t#gK@iNC)l#nvH݁TQ>A*ٓ3ۙ @D׸K oU7OyJCҸ5ǘ5UaH?nӫWÈ9PW(ǜ,n`(N15B]å#MيhSA^c+V}R]Īlnfz<*Ehyi66t0ihHUɃY磱L#me#Le=wo'4R(iwtVGNI?~Z|B>lt=:~̴eې*.ai}ކmX^~N\])mBrEr0*)uMaEҿPN>H f`_cKz zR?~4YĹ X`⎫4**1R-f3QR\QaS*mq49Wu$,:Ac{\. @;v܁\|jwP?pUZnKITl' ^x3qOzl+/ ߡ.9FPog3IC'tuoe5*JSNb~^Ȝ }jg裮0Pż6?NRg~8Ğs}>{W^'L=Oڝؚ>+qPȳX)oD[|۫T$wte n`?G5Po~9mU R"dpn83aS> ?5%}%.)p4$[zn $a)Ks,#w nF4pP@*ܑWpTjgVvﴠ#UFW R|ms3Й+z5,Q/93OOsˮ&P Aבxߢ 7XG@g`{{ۉ8|?E R'a8ƊaU=ٴpV-NTQ]qA?+|,@5+WLFWo?c} 6sEK\7b54=8G[ÈoI[wPis!sK|gsm*HE1YNy7J%< V@ڪUB9- Jaxa|3ho75 #Ev/ ^7 4^ݬӪpc{5RoJ|x@?!W{|eZU.O 2i 6Iu0Zs2%!c@й%63>@ߨ%Zc'BWC!z,(og{ wfcpgF_VP? }̧sB# aW\uH* .cJe5!OF (MP~Z ɦCTyGLF-Cs R<GVCիa //{zi/W 0y8) Ȃ^NNѧISҕQ0ZcՇZ͘#~\ݛ؛PlBTuP̛Xot(78C|bAǙ-Oi< Ja>5Rq>+M_ì#+/7I z[JF$n$Pv=!fN3FR̤4l ߺ/e7RD;`xV\< a%Fxz֠6X>L.y۫pEԋ_gv9Owq˽,=ߦto;|ޮktrVw羯)4B08"wJ}xuikRoLލ/HڶJ%ʙ ./(7 .(\GSw_\c7]X bjbDž$)xU? ߟe:ޘJEb[[,./kӮz41ٺ@\_r2WZ]od*׵=J)Ȣs¤w/ItN<6W0E.VP>]&{<ڕA?gX.+Lr{i^uVb;'=l.k}/,Lmkx.aa|; @c;jd"uR@{؇wt@r@&xY%Vgq;nEA2e0#u:8M?}@ t_0 nE(r30j8c#|S Mn\!lg^Pin{wq*>[0Ry'a^V].+kf~Ũ]Y~6!egGrl%Q'i(<ٗ[~exnaٯ`܍B + e4&F8h5b @7&t*LЧ*ۏwWSu[ioߗbj#} Y)"tkaY:0Vre}u&ĝԋ^UK=641sb%_$}P!Vio(oC2M7@-CbDGHH/`&ׇ_pz<F yT- !nxB)W+wԯßjM皛;Ud-<#&ߴ!T5Ev5:aJ*Fzد+m(⥲뻧p*2 s?*'QlJ؝sꨊ (}[J^1K> ^a"$oځɋ;hkFr3;hj7*G@ ւ}WAԛv!/ӶZ;tZ ^Y}){7\=-߁^as"۟x{E/Ga2Odk}`6)܏i:T;CU ? zh(.}Cpqh?+pkȡ  Tshڽ3WJ? >mmin,J5*@-FkD tGG΍wb}m!E8^ux7qlkga5:x:C[Ipt^@5E;*t,Ro& gݐ=9$woI?LJO;f[Ca?2Dc28r4x)~Fups$z$H1;c,h!n,p 10,svXMFZ#V'P5 ||xWfr111/4'~N+~B=w0JU|wf2kg6iLvd"t[wVFX`b$x4{6jVGDGD@DDDB"Z#њ-D%:]]%]%] .Dg":3YHHl,#$6&$LbSHfX#5X38-ӑXGI\K$qI$.ĥ8"qFg"qfg!Z#$>ǓHH|2O!E$D$B$AGbIBI' $!$$dB $!$I$I$jI$ƒ8OHb"IL"$1$HbI4DI4D IҒ$I%Iq$)$%DDIR I2,d$I&d&I%:KHrO!ב T%V,%=D!PQ4[A1m/2m"tɻ pGA.~!h&k -I'rxOrBJy#B"~}zcf~yñ څ >FxQVkx^r ]U&9^ \c2L. HF>GgH0??z?PbU>7ͅvvKa6ڀ!.藼\hOІ15QaޠfPP>,0y =释h˓O32lbgz?*.I#BWb|NE0 A!a=ɚ-|eê,}x~Vn^-,B~2_ }00cnYB$~]تtP\eB}-xxlZ.C ]a_O>> Q^C9TxpQ]G:8Z\86X^ ʋƃ][_ݽ89BUWC>όi}` 264@#@JmR&Xf< m8 9  `"`&R&Xf< m8   aHõqPopJ;_CZ(8 4+}a_3[ 5P;;Ի;ןTW ycm@K/RJp囕ȻNT-~oc? yAGjkq~mR@^:Ce+< CdtB^ĵ'(p 0`>nE3,xeWV 6{>-vTo8#)@8p29c @[: n(q)  p&<0@4#  `p<0`!@O,x`K+^x-w|Ml(( `?WT8I~FP f0@@O>}b 0`iLx` @&@ `:`0 Q \OO< \1dZxt]QkGv =7X|׻>E*,}հ8iX]yrO8a"$G_i};_ҿ]z'?k6Z !3~2}gSxfN ̟k-i~xZFO",q0@|xo3<>n,?=xo-W{lLn}{ѡBߨw?b?<੮ZpvWrwgAaMb퇧9rR\NLQ,&w ?_jz*ud?xmm#G5\\∇.suc={4<>uڧǼΪ<~+.gwxyhkކ%17}~ύ5qԇ#'vZoLXפK&r5 gN6mlԎ _54[fwtY5~uegLԃq;>?h ϟiҜ8UO1W_W~uL_?k/uO+୴[W=5ϧsR53};3vq_ۆG'ڵ\W^9ީsݿu5M,p- _?~߫G>``ŕ]=', }|GǬBc}6Y MA =ҁp/ݳnY!mJ#.*.v^zƚm=Ij|Azqmm[c|aO[v5_0!"/*ΠUϰӢM{Tg-H q]β*N}'?^gmH/6x.)oy7dݼ6ve?7zOdc]<ޗ]Khsf #oZzʸ`蝁!w9hzvsN˼q=Rn}7>p-/m{fբHr$b]Y-Gݶߦ_H,߷a|`sfþS/ cKeLlG]KVSOH{;$|}?vzI|ʱ&hz˽!Wfmiw$r3 zMw⑵e;/?Řg'XޜWR::{UcKDzY]r'!']gߔsGImf@+m{!}-owG^ʹnOAk?:8D}IowyZW%^Ŗs-g_%i+^8O<57{rie/YEaz=MG~A?߻_B ޴ڣէ"w߹!6om?]Isp?BǷO:禯n}}&/G>gK”K2G\iU1Ϳ|>S;OcLS,& Xihn}~C-YsJ+dnL#ʬ}{ھ~h >s\3'ORKfvݰ3Ewf]O?o<88b!M{F_ ^~j/|v Cg/A|5T=}՘ 9<;av_u?[[Ʃ=~"ri|流&웎c4sWu}Q*6޳aST*nC/VW2磿Fo_:r[.;jgFN_ٙ>+y.>r9yA{4?~u/O;[8WhB7 ^ϧxɎ#MV^z2#ϧ>兟zK 9ws<U^/䰂o^P66'Vo]:vG~cv=H/Scys/~|߇!u~ѹb6=<\~{jw /7.G5=-O o;n3>wK7>Ru]gOE,|Mm7ܩ3<X>Mͧv+ Q,jVwYjjo#vE*J61"Y{ AK9ys.߻K?6)hOZZJ/`dR~7<ඨ1b?eۥ:k~闌 nv} &˷K<11Oޣe4}_!qB qW`6/zLZ]s,ek%; zdaK{EK/JPoDU(u.^Սr3lT{me_zqs2P"͵O7sBwN=u3S@OZz0pOCsfM}Uꃎ!%X68@A!/N0 %كZF5!ÂTc˹lQY (Rm2 Xwe$VOav0+*t{޻"R ;_R,=bx%܄P!+uoRq/}a's,&\J&T5`Qsz_DÄ7K b2t/S}:BV~ѹ0$Mz82ی zǪh)\EkIJS ~A @

Lqиļl2̒td񲽍|3-osG٘,};þB}饼w5 bFV"i']KDd ǃpyP*NpjrqgJ)xR,2&oog5G@fy[(au,(rVaL#H> JmTycUlóCGr0: PK_5fRޙEj,-mme]HIb<▦ozzcy4M8=0܎-#/ŇVk_`^O3ZbvMo>(:@?=Xz?MΙNHUM $k]@yD`Psl/lB>Y\ȍKzռEMaxT9h`}$NMe2"(,ӣjIb KAΫL`jXhi1mqzoyʬ&ы'0"|K˓r >nx{`-tgwm s(֍'wP6wF,^?c_lArS[Ӫ咳E ןO9IMMt#eJ#x2͸=_7wfsK ..LE弊4PYI`~ 7)Hٝz^g.D;{H"!OemEX`QNP=t4Eׅ5Vv>_'m+? u \ejc/VMMPˬ:>8d˃Th3PQ['-}2=}q|9+ގQdՋ1'$VPJZ&r3s$W\qQ+=cd|qXzO+t,#~epcP3M>3s(B! Dp|Ԙ0j,x96V[h$}-!q@D !{k1"]]HmI;uW\d`4 {uvUzwW CYY&qo?Is~=g<VNjnUAqǯ}@+h.) "!ԚSM:!CF9l\ԷS8ӭ|y>ٿ}a=4~ܼSAwڰ5w-ۂw|R@e*N۠GT3Y"K0BUw!n7󉴰ӝۖ;>{"tVrNتt<>RZAĵ3l1SJطg+Shӥ{*;-]C߳:DF tpi|/k]dS!?-4#{(#"$ yVk$]IPc>lF|WvG.˷<7~'2uRܘ'P+7sM(<['M0o{¾^'o&Ȗ5f XƧyTE3xhfG-78އ]ys&_O^ &4SEgvyn~όIv 2@4{ECV_(b(*jTb(Ե4׾K:"\yv>~ E7l}'<  V5U"F3Xv&;(3Թj=`>.T̉SSvu x +~jNS'H=WN3~#mcZ#8#x@}ّOR7Wkoy?Okz-r*u8*H^\ ٱT&  ^Bf*+>/Ƃ PJ籋]975&&!Uh[pζV*饜1uZSEܳ?Z&BkatLOS]z ǁG º&3OHR<9.-+v(ـf~ro3ZWCB{:W$\(Dۍe*kkX4{33TU9e ǧ~FHM:&ל/ށDʆt g8:/oNQ!mYJvG& ;x9P"оs>]Ez0Uj9uIމ%rI"Jc}M$vg{c"67UU ' Qsij=yiUZljc D4f!c$"C!goŗo4-fIzao苦M59,>[N|.EI\.vOjQ=-^@ekӁj7\B5cI&ێYKG&NFҪc ^d?qk_ޜ{6zn›^