Bug#1078229: xserver-xorg-video-ivtvdev: FTBFS with GCC 14: error: assignment ... from incompatible pointer type ...
Source: xserver-xorg-video-ivtvdev Version: 1.1.2-2 Severity: serious Tags: ftbfs sid trixie Justification: fails to build from source (but built successfully in the past) Hi, xserver-xorg-video-ivtvdev started to FTBFS when GCC 14 was made the default compiler: Making all in src make[3]: Entering directory '/build/xserver-xorg-video-ivtvdev-1.1.2/build/src' /bin/bash ../libtool --tag=CC --mode=compile gcc -DHAVE_CONFIG_H -I. -I../../src -I..-I/usr/include/xorg -fvisibility=hidden -I/usr/include/pixman-1 -I/usr/include/X11/dri -I/usr/include/libdrm -g -O2 -c -o ivtv.lo ../../src/ivtv.c libtool: compile: gcc -DHAVE_CONFIG_H -I. -I../../src -I.. -I/usr/include/xorg -fvisibility=hidden -I/usr/include/pixman-1 -I/usr/include/X11/dri -I/usr/include/libdrm -g -O2 -c ../../src/ivtv.c -fPIC -DPIC -o .libs/ivtv.o ../../src/ivtv.c: In function 'IVTVDevProbe': ../../src/ivtv.c:350:13: warning: assignment discards 'const' qualifier from pointer target type [-Wdiscarded-qualifiers] 350 | dev = xf86FindOptionValue(devSections[i]->options, "fbdev"); | ^ ../../src/ivtv.c:373:26: error: assignment to 'ModeStatus (*)(struct _ScrnInfoRec *, struct _DisplayModeRec *, Bool, int)' {aka 'ModeStatus (*)(struct _ScrnInfoRec *, struct _DisplayModeRec *, int, int)'} from incompatible pointer type 'int (*)(int, struct _DisplayModeRec *, Bool, int)' {aka 'int (*)(int, struct _DisplayModeRec *, int, int)'} [-Wincompatible-pointer-types] 373 | pScrn->ValidMode = ivtvHWValidMode; | ^ ../../src/ivtv.c: In function 'IVTVDevPreInit': ../../src/ivtv.c:423:13: warning: passing argument 3 of 'ivtvHWInit' discards 'const' qualifier from pointer target type [-Wdiscarded-qualifiers] 423 | xf86FindOptionValue(devPtr->pEnt->device->options, "fbdev"))) | ^~~ In file included from ../../src/ivtv.c:36: ../../src/ivtv_hw.h:65:70: note: expected 'char *' but argument is of type 'const char *' 65 | Bool ivtvHWInit(ScrnInfoPtr pScrn, struct pci_device *PciInfo, char *device); | ~~^~ make[3]: *** [Makefile:338: ivtv.lo] Error 1 make[3]: Leaving directory '/build/xserver-xorg-video-ivtvdev-1.1.2/build/src' Andreas xserver-xorg-video-ivtvdev_1.1.2-2.log.gz Description: application/gzip
Bug#1069405: x11-apps: FTBFS on arm64: configure: error: Package requirements (x11-xcb xcb-present >= 1.9 xcb-xfixes xcb-damage) were not met
Followup-For: Bug #1069405 Control: retitle -1 x11-apps: FTBFS with libxcb 1.17: Package requirements (x11-xcb xcb-present >= 1.9 xcb-xfixes xcb-damage) were not met Control: block -1 with 1069408 This happens on all architectures and is probably caused by the missing libxcb-dri3-dev dependency in libxcb-present-dev. (#1069408) Andreas
Bug#1067012: renderdoc: FTBFS with glslang 14: fatal error: glslang/Include/Types.h: No such file or directory
Source: renderdoc Version: 1.27+dfsg-1 Severity: serious Tags: ftbfs Justification: fails to build from source (but built successfully in the past) renderdoc FTBFS with glslang 14 in sid due to a missing header: [184/311] /usr/bin/c++ -DAMD_EXTENSIONS -DDISTRIBUTION_CONTACT=\"https://salsa.debian.org/xorg-team/app/renderdoc\; -DDISTRIBUTION_NAME=\"Debian\" -DDISTRIBUTION_VERSION=\"1.27+dfsg-1\" -DNV_EXTENSIONS -DRELEASE -DRENDERDOC_EXPORTS -DRENDERDOC_PLATFORM_LINUX -DRENDERDOC_STABLE_BUILD=1 -DRENDERDOC_SUPPORT_EGL -DRENDERDOC_SUPPORT_GL -DRENDERDOC_SUPPORT_GLES -DRENDERDOC_SUPPORT_VULKAN -DRENDERDOC_VULKAN_JSON_SUFFIX="" -DRENDERDOC_WINDOWING_XCB -DRENDERDOC_WINDOWING_XLIB -DRENDERDOC_X86_PROC_FAMILY=1 -I/build/rende rdoc-1.27+dfsg/renderdoc -I/build/renderdoc-1.27+dfsg/renderdoc/3rdparty -g -O2 -ffile-prefix-map=/build/renderdoc-1.27+dfsg=. -fstack-protector-strong -fstack-clash-protection -Wformat -Werror=format-security -fcf-protection -Wdate-time -D_FORTIFY_SOURCE=2 -std=c++11 -fno-strict-aliasing -fvisibility=hidden -fvisibility-inlines-hidden -Wall -Wextra -Wno-unused-variable -Wno-unused-parameter -Wno-unused-result -Wno-type-limits -Wno-missing-field-initializers -Wno-unknown-pragmas -Wno-reorder -Wno-unused-but-set -variable -Wno-maybe-uninitialized -Wno-class-memaccess -Wimplicit-fallthrough=2 -Wno-unused-value -O3 -DNDEBUG -fPIC -MD -MT renderdoc/driver/shaders/spirv/CMakeFiles/rdoc_spirv.dir/glslang_compile.cpp.o -MF renderdoc/driver/shaders/spirv/CMakeFiles/rdoc_sp irv.dir/glslang_compile.cpp.o.d -o renderdoc/driver/shaders/spirv/CMakeFiles/rdoc_spirv.dir/glslang_compile.cpp.o -c /build/renderdoc-1.27+dfsg/renderdoc/driver/shaders/spirv/glslang_compile.cpp FAILED: renderdoc/driver/shaders/spirv/CMakeFiles/rdoc_spirv.dir/glslang_compile.cpp.o /usr/bin/c++ -DAMD_EXTENSIONS -DDISTRIBUTION_CONTACT=\"https://salsa.debian.org/xorg-team/app/renderdoc\; -DDISTRIBUTION_NAME=\"Debian\" -DDISTRIBUTION_VERSION=\"1.27+dfsg-1\" -DNV_EXTENSIONS -DRELEASE -DRENDERDOC_EXPORTS -DRENDERDOC_PLATFORM_LINUX -DRENDERD OC_STABLE_BUILD=1 -DRENDERDOC_SUPPORT_EGL -DRENDERDOC_SUPPORT_GL -DRENDERDOC_SUPPORT_GLES -DRENDERDOC_SUPPORT_VULKAN -DRENDERDOC_VULKAN_JSON_SUFFIX="" -DRENDERDOC_WINDOWING_XCB -DRENDERDOC_WINDOWING_XLIB -DRENDERDOC_X86_PROC_FAMILY=1 -I/build/renderdoc-1.27+ dfsg/renderdoc -I/build/renderdoc-1.27+dfsg/renderdoc/3rdparty -g -O2 -ffile-prefix-map=/build/renderdoc-1.27+dfsg=. -fstack-protector-strong -fstack-clash-protection -Wformat -Werror=format-security -fcf-protection -Wdate-time -D_FORTIFY_SOURCE=2 -std=c++11 -fno-strict-aliasing -fvisibility=hidden -fvisibility-inlines-hidden -Wall -Wextra -Wno-unused-variable -Wno-unused-parameter -Wno-unused-result -Wno-type-limits -Wno-missing-field-initializers -Wno-unknown-pragmas -Wno-reorder -Wno-unused-but-set-variable -Wno-maybe-uninitialized -Wno-class-memaccess -Wimplicit-fallthrough=2 -Wno-unused-value -O3 -DNDEBUG -fPIC -MD -MT renderdoc/driver/shaders/spirv/CMakeFiles/rdoc_spirv.dir/glslang_compile.cpp.o -MF renderdoc/driver/shaders/spirv/CMakeFiles/rdoc_spirv.dir/gl slang_compile.cpp.o.d -o renderdoc/driver/shaders/spirv/CMakeFiles/rdoc_spirv.dir/glslang_compile.cpp.o -c /build/renderdoc-1.27+dfsg/renderdoc/driver/shaders/spirv/glslang_compile.cpp /build/renderdoc-1.27+dfsg/renderdoc/driver/shaders/spirv/glslang_compile.cpp:32:10: fatal error: glslang/Include/Types.h: No such file or directory 32 | #include "glslang/Include/Types.h" | ^ compilation terminated. Andreas
Bug#1065551: spirv-headers: please update to commit b73e168ca5e123dcf3dea8a34b19a5130f421ae1 or later
Source: spirv-headers Version: 1.6.1+1.3.275.0-1 Severity: wishlist Hi, spirv-llvm-translator upstream just had in its llvm_release_180 branch a commit (d970c9126c033ebcbb7187bc705eae2e54726b74) pushed that bumps the header dependency to b73e168ca5e123dcf3dea8a34b19a5130f421ae1 for using SPV_INTEL_maximum_registers. This will be part of the 18.0.0 release soon and in order to package that we need some newer spirv-headers packaged, too. AFAICS there is no new spirv-headers tag yet that includes this commit. Thanks Andreas
Bug#1053796: spirv-headers: please update to sdk-1.3.261.0 or later
Package: spirv-headers Version: 1.6.1+1.3.250.0-1 Severity: serious Hi, please update spirv-headers to a newer release. For creating src:spirv-llvm-translator-17 we need at least commit 9b527c0fb60124936d0906d44803bec51a0200fb aka sdk-1.3.261.0~10^2~1 Thanks Andreas
Bug#1040056: [Pkg-opencl-devel] Bug#1040056: spirv-tools breaks spirv-llvm-translator-15 autopkgtest: exactly one input file must be specified.
Does this issue affect spirv-llvm-translator-16, too? I tried to trigger the corresponding tests (which succeeded), but I'm not sure if it really tested the correct setup. Andreas
Bug#970556: vulkan-validationlayers-dev: ships /usr/include/xxhash.h already provided by libxxhash-dev
Package: vulkan-validationlayers-dev Version: 1.2.148.0-1 Severity: serious User: debian...@lists.debian.org Usertags: piuparts Hi, during a test with piuparts I noticed your package failed to install because it tries to overwrite other packages files. >From the attached log (scroll to the bottom...): Preparing to unpack .../vulkan-validationlayers-dev_1.2.148.0-1_amd64.deb ... Unpacking vulkan-validationlayers-dev:amd64 (1.2.148.0-1) ... dpkg: error processing archive /var/cache/apt/archives/vulkan-validationlayers-dev_1.2.148.0-1_amd64.deb (--unpack): trying to overwrite '/usr/include/xxhash.h', which is also in package libxxhash-dev:amd64 0.8.0-1 Errors were encountered while processing: /var/cache/apt/archives/vulkan-validationlayers-dev_1.2.148.0-1_amd64.deb cheers, Andreas libxxhash-dev=0.8.0-1_vulkan-validationlayers-dev=1.2.148.0-1.log.gz Description: application/gzip
Bug#970134: libweston-9-dev: missing (unversioned) Breaks+Replaces: libweston-8-dev
Package: libweston-9-dev Version: 9.0.0-1 Severity: serious User: debian...@lists.debian.org Usertags: piuparts Hi, during a test with piuparts I noticed your package fails to upgrade from 'testing'. It installed fine in 'testing', then the upgrade to 'sid' fails because it tries to overwrite other packages files without declaring a Breaks+Replaces relation. See policy 7.6 at https://www.debian.org/doc/debian-policy/ch-relationships.html#overwriting-files-and-replacing-packages-replaces >From the attached log (scroll to the bottom...): Preparing to unpack .../libweston-9-dev_9.0.0-1_amd64.deb ... Unpacking libweston-9-dev (9.0.0-1) ... dpkg: error processing archive /var/cache/apt/archives/libweston-9-dev_9.0.0-1_amd64.deb (--unpack): trying to overwrite '/usr/lib/x86_64-linux-gnu/weston/libexec_weston.so', which is also in package libweston-8-dev 8.0.0-1 Errors were encountered while processing: /var/cache/apt/archives/libweston-9-dev_9.0.0-1_amd64.deb Since libweston-8-dev is gone, the Breaks+Replaces can be unversioned. cheers, Andreas libweston-8-dev=8.0.0-1_libweston-9-dev=9.0.0-1.log.gz Description: application/gzip
Bug#943664: xserver-xorg-video-radeon 19.1 (ppc64) crash at startup: undefined symbol: exaGetPixmapDriverPrivate
Control: severity -1 important On Sun, 27 Oct 2019 16:59:23 +0100 "Ralf P." wrote: > Package: xserver-xorg-video-radeon > Version: 1:19.0.1-1 > Severity: grave > Justification: renders package unusable > > Dear Maintainer, > >on the ppc64 architecture the package Downgrading the severity for a Debian ports architecture. Andreas
Bug#947392: mesa-common-dev: missing Depends: libgl-dev (>= 1.3)
Package: mesa-common-dev Version: 19.3.0~rc6-1 Severity: serious mesa-common-dev needs to depend on libgl-dev after the headers were moved. The missing headers cause e.g. a FTBFS in pycuda. Andreas
Bug#929526: xfonts-scalable: fails to install in lenny/i386: fmt: invalid width: `63482'
Package: xfonts-scalable Version: 1:1.0.0-6 Severity: serious User: debian...@lists.debian.org Usertags: piuparts Hi, during a test with piuparts I noticed your package failed to install. As per definition of the release team this makes the package too buggy for a release, thus the severity. >From the attached log (scroll to the bottom...): Setting up xfonts-scalable (1:1.0.0-6) ... fmt: invalid width: `63482' fmt: invalid width: `63482' fmt: invalid width: `63482' dpkg: error processing xfonts-scalable (--configure): subprocess post-installation script returned error exit status 1 Errors were encountered while processing: xfonts-scalable It installs fine in squeeze/i386 with the version from squeeze. cheers, Andreas xfonts-scalable_1:1.0.0-6.log.gz Description: application/gzip
Bug#926857: mesa-vdpau-drivers: broken symlink: /usr/lib/x86_64-linux-gnu/vdpau/libvdpau_gallium.so -> libvdpau_gallium.so.1.0.0
Package: mesa-vdpau-drivers Version: 18.3.6-1 Severity: serious User: debian...@lists.debian.org Usertags: piuparts Hi, during a test with piuparts I noticed your package ships (or creates) a broken symlink. >From the attached log (scroll to the bottom...): 0m23.0s ERROR: FAIL: Broken symlinks: /usr/lib/x86_64-linux-gnu/vdpau/libvdpau_gallium.so -> libvdpau_gallium.so.1.0.0 (mesa-vdpau-drivers:amd64) cheers, Andreas mesa-vdpau-drivers_18.3.6-1.log.gz Description: application/gzip
Bug#903929: libglx-mesa0: please add Breaks: glx-diversions (<< 0.8.4~)
Package: libglx-mesa0 Version: 18.1.4-1 Severity: normal Hi, please add a Breaks: glx-diversions (<< 0.8.4~) to libglx-mesa0. The older versions did not handle libGLX_indirect.so.0 which could lead to the nvidia driver removing the libGLX_indirect.so.0 symlink. Thanks Andreas
Bug#896602: libglvnd: FTBFS on many architectures
Source: libglvnd Version: 1.0.0+git20180308-1 Severity: serious Justification: fails to build from source (but built successfully in the past) Hi, the snapshot uploaded to sid FTBFS on many architectures where it previously built. https://buildd.debian.org/status/package.php?p=libglvnd=unstable Andreas
Bug#884717: libgl1: consider using a version that is higher than 1.2.0 for the libGL.so.1.0.0 filename
Control: tag -1 fixed-upstream On 2018-01-25 17:22, Timo Aaltonen wrote: >> Also give me a notice before uploading such a change s.t. I can adjust >> glx-diversions accordingly and you can bump the Breaks against the >> glx-diversions versions not knowing about the new filename. >> >> It may be worthwile suggesting such a rename to upstream, too. > > Correction, it should go upstream _first_. Upstream will bump all the filename versions to prevent accidently mixing libglvnd and mesa libraries ... https://github.com/NVIDIA/libglvnd/commit/212861e5e73927ec267244f5fc3040bb9b701c47 Please bump the Breaks against glx-diversions to (<< 0.8.3~) (I just uploaded 0.8.3 to sid). Thanks
Bug#884717: libgl1: consider using a version that is higher than 1.2.0 for the libGL.so.1.0.0 filename
Package: libgl1 Version: 1.0.0-1 Severity: normal Hi, if the system got messed up by some proprietary installer, it may be well possible that some cruft libGL.so.1.X.Y outside the control of dpkg is left on the system and takes precedence over libGL.so.1.0.0 Like in #879041: libglvnd0: Nvidia-Installer 384.90: "An incomplete installation of libglvnd was found." where an old copy of the MESA libGL.so.1.2.0 was left and was still ebing used. ldconfig will give precedence to the file with the highest filename version as the target of the SONAME link libGL.so.1. So unfortunately libGL.so.1.0.0 will always be the loser. Renaming libGL.so.1.0.0 (or just adding a symlink with a higher version to it) would circumvent these bugs by giving precedence to the libglvnd libGL.so.1. The following filenames have been in use in the past and present: libgl1: /usr/lib/x86_64-linux-gnu/libGL.so.1.0.0 libgl1-fglrx-glx: /usr/lib/x86_64-linux-gnu/fglrx/fglrx-libGL.so.1.2 libgl1-fglrx-legacy-glx: /usr/lib/x86_64-linux-gnu/fglrx/fglrx-libGL.so.1.2 libgl1-glvnd-nvidia-glx: /usr/lib/x86_64-linux-gnu/nvidia/current/libGL.so.1.0.0 libgl1-mesa-glx: /usr/lib/x86_64-linux-gnu/libGL.so.1.2 libgl1-mesa-glx: /usr/lib/x86_64-linux-gnu/libGL.so.1.2.0 libgl1-mesa-swx11: /usr/lib/x86_64-linux-gnu/libGL.so.1.5.08005 libgl1-mesa-swx11: /usr/lib/x86_64-linux-gnu/libGL.so.1.6.0 I would suggest to add a symlink libGL.so.1.7.0 -> libGL.so.1.0.0 That also means to adjust the SONAME link to libGL.so.1 -> libGL.so.1.7.0 to ship the link like it would be created by ldconfig, otherwise it may trigger errors in some library symlink validation test in piuparts. Also give me a notice before uploading such a change s.t. I can adjust glx-diversions accordingly and you can bump the Breaks against the glx-diversions versions not knowing about the new filename. It may be worthwile suggesting such a rename to upstream, too. Andreas
Bug#872874: libxkbcommon-x11-dev: Missing depends on libxkbcommon-dev
Followup-For: Bug #872874 Fixed in stretch with the attached patch. Andreas diff -u libxkbcommon-0.7.1/debian/changelog libxkbcommon-0.7.1/debian/changelog --- libxkbcommon-0.7.1/debian/changelog +++ libxkbcommon-0.7.1/debian/changelog @@ -1,3 +1,18 @@ +libxkbcommon (0.7.1-2~deb9u1) stretch; urgency=medium + + * Non-maintainer upload. + * Rebuild for stretch. + + -- Andreas Beckmann <a...@debian.org> Mon, 27 Nov 2017 17:50:43 +0100 + +libxkbcommon (0.7.1-2) unstable; urgency=medium + + * Remove Cyril from Uploaders. + * Add missing dependency libxkbcommon-x11-dev → libxkbcommon-dev +(closes: #872874). + + -- Julien Cristau <jcris...@debian.org> Sat, 16 Sep 2017 13:40:36 +0200 + libxkbcommon (0.7.1-1) unstable; urgency=medium * New upstream release. diff -u libxkbcommon-0.7.1/debian/control libxkbcommon-0.7.1/debian/control --- libxkbcommon-0.7.1/debian/control +++ libxkbcommon-0.7.1/debian/control @@ -2,7 +2,7 @@ Section: x11 Priority: optional Maintainer: Debian X Strike Force <debian-x@lists.debian.org> -Uploaders: Cyril Brulebois <k...@debian.org>, Michael Stapelberg <stapelb...@debian.org> +Uploaders: Michael Stapelberg <stapelb...@debian.org> Build-Depends: debhelper (>= 10), quilt, @@ -94,6 +94,7 @@ Pre-Depends: ${misc:Pre-Depends} Depends: libxkbcommon-x11-0 (= ${binary:Version}), + libxkbcommon-dev (= ${binary:Version}), libxcb1-dev, libxcb-xkb-dev, ${shlibs:Depends},
Bug#882682: libglvnd: make symbols files generate versioned dependencies
Source: libglvnd Version: 1.0.0-1 Severity: wishlist Hi, I'd like to see all the .symbols files in libglvnd generating versioned dependencies. That will be achieved by setting all the symbol versions to something higher than 0. This would mean: * dependencies are no longer satisfiable by the virtual packages from src:mesa (<< 17) (a package built on buster won't be installable on stretch) * it would simplify my work matching src:libglvnd package versions with src:nvidia-graphics-drivers packages, should I ever consider providing libglvnd package names again (since I could provide a version that is *less* than any src:libglvnd version, which doesn't work if the dependency other packages having on e.g. libgl1 is unversioned) If you agree, I can prepare a patch. Andreas
Bug#879041: libglvnd0: Nvidia-Installer 384.90: "An incomplete installation of libglvnd was found."
On Wed, 18 Oct 2017 19:07:41 +0200 Dirk Lehmannwrote: > During installing the Nvidia Geforce driver version 384.90 the > following error occurs: Please use the 384.xx nvidia-graphics-drivers packages from experimental. Andreas
Bug#876084: [libglvnd-dev] Can't upgrade libglvnd-dev anymore
Followup-For: Bug #876084 This bug could be related to several similar ones against nvidia-graphics-drivers recently, finally fixed with 375.82-8 ... I suggest to add a bug-control file to automatically record this fact in case of bug reports. libgl1-nvidia-glx-any is a virtual package that is provided by src:nvidia-graphics-drivers for several years already. And there is another little patch switching to dh_missing ... The --builddirectory=build/ is a bit unfortunate, since it clashes with 'dh build' in case someone manually runs it twice - which I obviously did :-) Andreas >From 3e268f4e30daf6fc78ba93f0ed134103da40006d Mon Sep 17 00:00:00 2001 From: Andreas Beckmann <a...@debian.org> Date: Sat, 25 Nov 2017 16:02:17 + Subject: [PATCH 3/4] bug-control: report whether the proprietary nvidia driver is installed --- debian/bug-control | 1 + debian/changelog | 2 ++ debian/rules | 3 +++ 3 files changed, 6 insertions(+) create mode 100644 debian/bug-control diff --git a/debian/bug-control b/debian/bug-control new file mode 100644 index 000..584c4f9 --- /dev/null +++ b/debian/bug-control @@ -0,0 +1 @@ +package-status: libgl1-nvidia-glx-any diff --git a/debian/changelog b/debian/changelog index 76f7e56..a721dfc 100644 --- a/debian/changelog +++ b/debian/changelog @@ -3,6 +3,8 @@ libglvnd (1.0.0-2) UNRELEASED; urgency=medium * Use source format 3.0 (quilt). * kfreebsd-hurd.patch: New, fix FTBFS on kFreeBSD and Hurd. (Closes: #870445) + * Add bug-control file to report whether the proprietary nvidia driver is +installed. -- Andreas Beckmann <a...@debian.org> Sat, 25 Nov 2017 14:59:46 + diff --git a/debian/rules b/debian/rules index 4f302d8..e3df5b3 100755 --- a/debian/rules +++ b/debian/rules @@ -17,6 +17,9 @@ override_dh_auto_test: override_dh_makeshlibs: dh_makeshlibs -a -- -c4 +override_dh_bugfiles: + dh_bugfiles -A + %: dh $@ --builddirectory=build/ -- 2.5.1 >From 2b6d0c464c7c346babcb0bb20d38a238b6d51dfc Mon Sep 17 00:00:00 2001 From: Andreas Beckmann <a...@debian.org> Date: Sat, 25 Nov 2017 16:05:28 + Subject: [PATCH 4/4] switch to dh_missing --fail-missing --- debian/changelog | 1 + debian/rules | 6 -- 2 files changed, 5 insertions(+), 2 deletions(-) diff --git a/debian/changelog b/debian/changelog index a721dfc..df4ac1f 100644 --- a/debian/changelog +++ b/debian/changelog @@ -5,6 +5,7 @@ libglvnd (1.0.0-2) UNRELEASED; urgency=medium (Closes: #870445) * Add bug-control file to report whether the proprietary nvidia driver is installed. + * Switch to dh_missing --fail-missing. -- Andreas Beckmann <a...@debian.org> Sat, 25 Nov 2017 14:59:46 + diff --git a/debian/rules b/debian/rules index e3df5b3..e35daf7 100755 --- a/debian/rules +++ b/debian/rules @@ -3,13 +3,15 @@ # Uncomment this to turn on verbose mode. #export DH_VERBOSE=1 -override_dh_install: +override_dh_auto_install: + dh_auto_install find debian/tmp -name '*.la' -delete # drop GLESv1 rm -f debian/tmp/usr/lib/*/libGLESv1* - dh_install --fail-missing +override_dh_missing: + dh_missing --fail-missing # needs X override_dh_auto_test: -- 2.5.1
Bug#870445: libglvnd: FTBFS on non-Linux: u_execmem_free undefined
Followup-For: Bug #870445 Control: tag -1 patch Hi, attached are two patches to fix this FTBFS. Build-tested on falla and exodar. The git repository is missing the pristine-tar commit for 1.0.0 and all the upstream version tags. Andreas >From 09a59c852e307c7e4d6770fa137975dab93ce9fe Mon Sep 17 00:00:00 2001 From: Andreas Beckmann <a...@debian.org> Date: Sat, 25 Nov 2017 15:00:32 + Subject: [PATCH 1/2] use source format 3.0 (quilt) --- debian/changelog | 6 ++ debian/source/format | 2 +- 2 files changed, 7 insertions(+), 1 deletion(-) diff --git a/debian/changelog b/debian/changelog index 207a54a..ded1623 100644 --- a/debian/changelog +++ b/debian/changelog @@ -1,3 +1,9 @@ +libglvnd (1.0.0-2) UNRELEASED; urgency=medium + + * Use source format 3.0 (quilt). + + -- Andreas Beckmann <a...@debian.org> Sat, 25 Nov 2017 14:59:46 + + libglvnd (1.0.0-1) unstable; urgency=high [ Andreas Boll ] diff --git a/debian/source/format b/debian/source/format index d3827e7..163aaf8 100644 --- a/debian/source/format +++ b/debian/source/format @@ -1 +1 @@ -1.0 +3.0 (quilt) -- 2.1.4 >From b48012726befb68920af1cbe4fc3beb0bc6afab2 Mon Sep 17 00:00:00 2001 From: Andreas Beckmann <a...@debian.org> Date: Sat, 25 Nov 2017 15:10:20 + Subject: [PATCH 2/2] fix FTBFS on kFreeBSD and Hurd --- debian/changelog | 2 ++ debian/patches/kfreebsd-hurd.patch | 14 ++ debian/patches/series | 1 + 3 files changed, 17 insertions(+) create mode 100644 debian/patches/kfreebsd-hurd.patch create mode 100644 debian/patches/series diff --git a/debian/changelog b/debian/changelog index ded1623..76f7e56 100644 --- a/debian/changelog +++ b/debian/changelog @@ -1,6 +1,8 @@ libglvnd (1.0.0-2) UNRELEASED; urgency=medium * Use source format 3.0 (quilt). + * kfreebsd-hurd.patch: New, fix FTBFS on kFreeBSD and Hurd. +(Closes: #870445) -- Andreas Beckmann <a...@debian.org> Sat, 25 Nov 2017 14:59:46 + diff --git a/debian/patches/kfreebsd-hurd.patch b/debian/patches/kfreebsd-hurd.patch new file mode 100644 index 000..c504e51 --- /dev/null +++ b/debian/patches/kfreebsd-hurd.patch @@ -0,0 +1,14 @@ +Author: Andreas Beckmann <a...@debian.org> +Description: fix FTBFS on kFreeBSD and Hurd + +--- a/src/GLdispatch/vnd-glapi/u_execmem.c b/src/GLdispatch/vnd-glapi/u_execmem.c +@@ -68,7 +68,7 @@ static unsigned char *exec_mem = NULL; + static unsigned char *write_mem = NULL; + + +-#if defined(__linux__) || defined(__OpenBSD__) || defined(_NetBSD__) || defined(__sun) || defined(__HAIKU__) ++#if defined(__linux__) || defined(__OpenBSD__) || defined(_NetBSD__) || defined(__sun) || defined(__HAIKU__) || defined(__FreeBSD_kernel__) || defined(__GNU__) + + #include + #include diff --git a/debian/patches/series b/debian/patches/series new file mode 100644 index 000..1dfcb87 --- /dev/null +++ b/debian/patches/series @@ -0,0 +1 @@ +kfreebsd-hurd.patch -- 2.1.4
Bug#876100: fixed in nvidia-graphics-drivers 375.82-4
On 20.09.2017 20:39, Jiri Palecek wrote: >> * Prevent mixing libgl1-nvidia-glx with libgl1-nvidia-glvnd-glx. >> * Use versioned Provides/Breaks/Replaces on the packages also >> built from >> src:libglvnd s.t. they cannot be satisfied by virtual packages >> provided >> from src:mesa (<< 17). (Closes: #875683, #876100) > I don't understand. How does this solve issue 876100? The problem was > that nvidia wasn't coinstallable with eg. libegl1-mesa-dev in its > current version, and it is still so. Could you explain? I just tried again in a minimal sid chroot and didn't encounter any problems: # apt-get install --install-recommends nvidia-driver [...] (successfully installed) # apt-get install libglvnd-dev Reading package lists... Done Building dependency tree Reading state information... Done The following additional packages will be installed: libglvnd-core-dev The following NEW packages will be installed: libglvnd-core-dev libglvnd-dev 0 upgraded, 2 newly installed, 0 to remove and 10 not upgraded. Need to get 0 B/16.3 kB of archives. After this operation, 80.9 kB of additional disk space will be used. Do you want to continue? [Y/n] [...] (successfully installed) # ls -lad $(dpkg -L libglvnd-dev) ls: cannot access 'diverted': No such file or directory ls: cannot access 'by': No such file or directory ls: cannot access 'glx-diversions': No such file or directory ls: cannot access 'to:': No such file or directory ls: cannot access 'diverted': No such file or directory ls: cannot access 'by': No such file or directory ls: cannot access 'glx-diversions': No such file or directory ls: cannot access 'to:': No such file or directory ls: cannot access 'diverted': No such file or directory ls: cannot access 'by': No such file or directory ls: cannot access 'glx-diversions': No such file or directory ls: cannot access 'to:': No such file or directory drwxr-xr-x 22 root root 440 Sep 20 22:51 /. drwxr-xr-x 10 root root 200 Mar 25 2013 /usr drwxr-xr-x 40 root root 880 Sep 20 22:54 /usr/lib lrwxrwxrwx 1 root root 15 Sep 12 07:32 /usr/lib/mesa-diverted/x86_64-linux-gnu/libEGL.so -> libEGL.so.1.0.0 lrwxrwxrwx 1 root root 14 Sep 12 07:32 /usr/lib/mesa-diverted/x86_64-linux-gnu/libGL.so -> libGL.so.1.0.0 lrwxrwxrwx 1 root root 18 Sep 12 07:32 /usr/lib/mesa-diverted/x86_64-linux-gnu/libGLESv2.so -> libGLESv2.so.2.0.0 drwxr-xr-x 26 root root 9300 Sep 20 22:56 /usr/lib/x86_64-linux-gnu lrwxrwxrwx 1 root root 49 Sep 20 22:56 /usr/lib/x86_64-linux-gnu/libEGL.so -> /etc/alternatives/glx--libEGL.so-x86_64-linux-gnu lrwxrwxrwx 1 root root 48 Sep 20 22:56 /usr/lib/x86_64-linux-gnu/libGL.so -> /etc/alternatives/glx--libGL.so-x86_64-linux-gnu lrwxrwxrwx 1 root root 52 Sep 20 22:56 /usr/lib/x86_64-linux-gnu/libGLESv2.so -> /etc/alternatives/glx--libGLESv2.so-x86_64-linux-gnu lrwxrwxrwx 1 root root 15 Sep 12 07:32 /usr/lib/x86_64-linux-gnu/libGLX.so -> libGLX.so.0.0.0 lrwxrwxrwx 1 root root 22 Sep 12 07:32 /usr/lib/x86_64-linux-gnu/libGLdispatch.so -> libGLdispatch.so.0.0.0 lrwxrwxrwx 1 root root 18 Sep 12 07:32 /usr/lib/x86_64-linux-gnu/libOpenGL.so -> libOpenGL.so.0.0.0 drwxr-xr-x 79 root root 1580 Sep 20 22:54 /usr/share drwxr-xr-x 397 root root 8640 Sep 20 22:56 /usr/share/doc drwxr-xr-x 2 root root 80 Sep 20 22:56 /usr/share/doc/libglvnd-dev -rw-r--r-- 1 root root 914 Sep 12 07:32 /usr/share/doc/libglvnd-dev/changelog.Debian.gz -rw-r--r-- 1 root root 4697 Sep 12 07:32 /usr/share/doc/libglvnd-dev/copyright # apt-get install libegl1-mesa-dev Reading package lists... Done Building dependency tree Reading state information... Done The following additional packages will be installed: libdrm-dev libpthread-stubs0-dev libwayland-bin libwayland-dev libx11-dev libx11-xcb-dev libxau-dev libxcb-dri2-0-dev libxcb-dri3-dev libxcb-glx0-dev libxcb-present-dev libxcb-randr0 libxcb-randr0-dev libxcb-render0-dev libxcb-shape0 libxcb-shape0-dev libxcb-sync-dev libxcb-xfixes0-dev libxcb1-dev libxdamage-dev libxdmcp-dev libxext-dev libxfixes-dev libxshmfence-dev libxxf86vm-dev x11proto-core-dev x11proto-damage-dev x11proto-dri2-dev x11proto-fixes-dev x11proto-gl-dev x11proto-input-dev x11proto-kb-dev x11proto-xext-dev x11proto-xf86vidmode-dev xorg-sgml-doctools xtrans-dev Suggested packages: libxcb-doc libxext-doc Recommended packages: libx11-doc The following NEW packages will be installed: libdrm-dev libegl1-mesa-dev libpthread-stubs0-dev libwayland-bin libwayland-dev libx11-dev libx11-xcb-dev libxau-dev libxcb-dri2-0-dev libxcb-dri3-dev libxcb-glx0-dev libxcb-present-dev libxcb-randr0 libxcb-randr0-dev libxcb-render0-dev libxcb-shape0 libxcb-shape0-dev libxcb-sync-dev libxcb-xfixes0-dev libxcb1-dev libxdamage-dev libxdmcp-dev libxext-dev libxfixes-dev libxshmfence-dev libxxf86vm-dev x11proto-core-dev x11proto-damage-dev x11proto-dri2-dev x11proto-fixes-dev x11proto-gl-dev x11proto-input-dev x11proto-kb-dev
Bug#876100: libglvnd-dev: libglvnd-dev not coinstallable with nvidia packages
On 18.09.2017 15:03, Jiri Palecek wrote: > libglvnd depends on many OpenGL related packages like libgl1, specified > by concrete version. This means those dependencies can't be satisfied > with virtual packages, ie. nvidia packages providing libgl1. However, > those nvidia packages conflict with any other packages providing those > names, particularly libgl1 et al. provided by libglvnd. First we have to sort out coexistence of the libglvnd runtime side of nvidia ... Andreas
Bug#870686: libglvnd: not ready for testing, yet
Source: libglvnd Version: 0.2.999+git20170802-2 Severity: serious Since the mesa packaging hasn't switched to glvnd, yet, there are file conflicts between several packages from src:mesa and src:libglvnd. So let's keep libglvnd out of testing until this transition progresses. Andreas
Bug#859929: x11-common: unowned files after upgrade from jessie to stretch and purge: /etc/X11/Xwrapper.config
On 2017-04-23 11:45, Julien Cristau wrote: > On Tue, Apr 11, 2017 at 22:20:59 -0400, G. Branden Robinson wrote: > >> At 2017-04-09T13:34:44+0200, Andreas Beckmann wrote: >>> 1m37.2s INFO: Warning: Package purging left files on system: >>> /etc/X11/Xwrapper.config not owned >> >> It appears the old Xwrapper was killed off at some point. >> >> The patch should be trivial enough that I can do it. This would also >> serve as an exercise of branden-guest's permissions on alioth. >> >> Any objection? >> > The wrapper was moved to the xserver-xorg-legacy package, and I'm a bit which is not installed by default during the upgrade ... > worried removing the configuration file in the x11-common maintainer > scripts would break that transition. Given that this is a complicated case ... I'm going to add the file to the ignore list built into piuparts. Andreas
Bug#859929: x11-common: unowned files after upgrade from jessie to stretch and purge: /etc/X11/Xwrapper.config
Package: x11-common Version: 1:7.7+18 Severity: important User: debian...@lists.debian.org Usertags: piuparts Hi, during a test with piuparts I noticed your package left unowned files on the system after purge, which is a violation of policy 6.8 (or 10.8): https://www.debian.org/doc/debian-policy/ch-maintainerscripts.html#s-removedetails Filing this as important as having a piuparts clean archive is a release goal since lenny. >From the attached log (scroll to the bottom...): 1m37.2s INFO: Warning: Package purging left files on system: /etc/X11/Xwrapper.config not owned This was observed after an upgrade from jessie to stretch. Works fine in both jessie and stretch without upgrading between releases. cheers, Andreas x11-common_1:7.7+18.log.gz Description: application/gzip
Bug#858469: xorg-docs: broken symlink: /usr/share/X11/doc -> ../doc/xorg-docs/docs
Package: xorg-docs Version: 1:1.7.1-1 Severity: normal User: debian...@lists.debian.org Usertags: piuparts Hi, during a test with piuparts I noticed your package ships (or creates) a broken symlink. >From the attached log (scroll to the bottom...): 0m25.2s ERROR: FAIL: Broken symlinks: /usr/share/X11/doc -> ../doc/xorg-docs/docs Maybe you wanted to target just ../doc/xorg-docs ? cheers, Andreas xorg-docs_1:1.7.1-1.log.gz Description: application/gzip
Bug#610693: Processed: reassign 446207 to src:tracker, reassign 610693 to src:xserver-xorg-video-openchrome ...
On 2016-07-24 10:33, Julien Cristau wrote: > On Sat, Jul 23, 2016 at 21:45:11 +, Debian Bug Tracking System wrote: > >>> reassign 610693 src:xserver-xorg-video-openchrome 1:0.2.904+svn842-2 >> Bug #610693 [xserver-xorg-video-via] xserver-xorg-video-via: >> KM400/KN400/P4M800 [S3 UniChrome] hangs KDE4 login >> Warning: Unknown package 'xserver-xorg-video-via' >> Bug reassigned from package 'xserver-xorg-video-via' to >> 'src:xserver-xorg-video-openchrome'. >> No longer marked as found in versions >> xserver-xorg-video-openchrome/1:0.2.904+svn842-2. The old "found" information referred to src:xserver-xorg-video-openchrome, so reassigning it back to the source package made sense to me. I assumed this was a transitional package in the old days and the maintainers might know how to deal with this bug properly. >> Ignoring request to alter fixed versions of bug #610693 to the same values >> previously set >> Bug #610693 [src:xserver-xorg-video-openchrome] xserver-xorg-video-via: >> KM400/KN400/P4M800 [S3 UniChrome] hangs KDE4 login >> Marked as found in versions xserver-xorg-video-openchrome/1:0.2.904+svn842-2. > > xserver-xorg-video-via and xserver-xorg-video-openchrome aren't the same > thing, I fail to see how reassigning this makes sense? Andreas PS: the list of orphaned bugs (filed against "packages without maintainers") left over from the removal of squeeze is now down from over 1000 to below 100
Bug#814631: xserver-xorg-video-siliconmotion: FTBFS: src/smi.h:224:5: error: unknown type name 'IOADDRESS'
Source: xserver-xorg-video-siliconmotion Version: 1:1.7.8-1 Severity: serious Hi, xserver-xorg-video-siliconmotion FTBFS in sid/amd64 (and probably other architectures as well), might be incompatible with recent Xorg Xserver: [...] dh_auto_build -O--builddirectory=build/ make -j1 make[1]: Entering directory '/build/xserver-xorg-video-siliconmotion-1.7.7/build' make all-recursive make[2]: Entering directory '/build/xserver-xorg-video-siliconmotion-1.7.7/build' Making all in src make[3]: Entering directory '/build/xserver-xorg-video-siliconmotion-1.7.7/build/src' /bin/bash ../libtool --tag=CC --mode=compile gcc -DHAVE_CONFIG_H -I. -I../../src -I..-fvisibility=hidden -I/usr/include/pixman-1 -I/usr/include/libdrm -I/usr/include/xorg -I/usr/include/X11/dri -Wall -Wpointer-arith -Wmissing-declarations -Wformat=2 -Wstrict-prototypes -Wmissing-prototypes -Wnested-externs -Wbad-function-cast -Wold-style-definition -Wdeclaration-after-statement -Wunused -Wuninitialized -Wshadow -Wmissing-noreturn -Wmissing-format-attribute -Wredundant-decls -Wlogical-op -Werror=implicit -Werror=nonnull -Werror=init-self -Werror=main -Werror=missing-braces -Werror=sequence-point -Werror=return-type -Werror=trigraphs -Werror=array-bounds -Werror=write-strings -Werror=address -Werror=int-to-pointer-cast -Werror=pointer-to-int-cast -fno-strict-aliasing -g -O2 -c -o smi_501.lo ../../src/smi_501.c libtool: compile: gcc -DHAVE_CONFIG_H -I. -I../../src -I.. -fvisibility=hidden -I/usr/include/pixman-1 -I/usr/include/libdrm -I/usr/include/xorg -I/usr/include/X11/dri -Wall -Wpointer-arith -Wmissing-declarations -Wformat=2 -Wstrict-prototypes -Wmissing-prototypes -Wnested-externs -Wbad-function-cast -Wold-style-definition -Wdeclaration-after-statement -Wunused -Wuninitialized -Wshadow -Wmissing-noreturn -Wmissing-format-attribute -Wredundant-decls -Wlogical-op -Werror=implicit -Werror=nonnull -Werror=init-self -Werror=main -Werror=missing-braces -Werror=sequence-point -Werror=return-type -Werror=trigraphs -Werror=array-bounds -Werror=write-strings -Werror=address -Werror=int-to-pointer-cast -Werror=pointer-to-int-cast -fno-strict-aliasing -g -O2 -c ../../src/smi_501.c -fPIC -DPIC -o .libs/smi_501.o In file included from ../../src/smi.h:40:0, from ../../src/smi_501.c:33: /usr/include/xorg/xf86PciInfo.h:50:2: warning: #warning "xf86PciInfo.h is deprecated. For greater compatibility, drivers should include necessary PCI IDs locally rather than relying on this file from xorg-server." [-Wcpp] #warning "xf86PciInfo.h is deprecated. For greater compatibility, drivers should include necessary PCI IDs locally rather than relying on this file from xorg-server." ^ In file included from ../../src/smi_501.c:33:0: ../../src/smi.h:224:5: error: unknown type name 'IOADDRESS' IOADDRESS PIOBase; /* Base of I/O ports */ ^ Makefile:521: recipe for target 'smi_501.lo' failed make[3]: *** [smi_501.lo] Error 1 make[3]: Leaving directory '/build/xserver-xorg-video-siliconmotion-1.7.7/build/src' Makefile:444: recipe for target 'all-recursive' failed make[2]: *** [all-recursive] Error 1 make[2]: Leaving directory '/build/xserver-xorg-video-siliconmotion-1.7.7/build' Makefile:376: recipe for target 'all' failed make[1]: *** [all] Error 2 make[1]: Leaving directory '/build/xserver-xorg-video-siliconmotion-1.7.7/build' dh_auto_build: make -j1 returned exit code 2 debian/rules:17: recipe for target 'build' failed make: *** [build] Error 2 dpkg-buildpackage: error: debian/rules build gave error exit status 2 Andreas
Bug#594269: Please reenable PCI_TXT_IDS_DIR for third-party drivers
Followup-For: Bug #594269 nvidia now implemented a different way to automatically integrate with Xorg via DRM (without needing an xorg.conf specifying a specific Driver), so this feature request is no longer relevant from my side. Andreas
Re: [Pkg-opencl-devel] New libclc snapshot
On 2015-09-03 14:57, Timo Aaltonen wrote: > On 02.09.2015 20:35, Andreas Boll wrote: >> Hi all, >> >> for mesa-opencl-icd we need a newer libclc snapshot which is >> compatible with LLVM-3.7, since we are upgrading mesa from LLVM-3.5 to >> 3.7. >> I'm volunteering to prepare such a new snapshot. >> So here is my question: >> Do you have for libclc a git repository for collaboration? >> If not could you create one under the umbrella of pkg-opencl? > I've put an updated package to > > git://git.debian.org/git/users/tjaalton/libclc.git > > feel free to use it as a base for the official update, and/or for the > migration to git. I imported the package history into git and rebased Timo's and Andreas' changes on top of this, the repository is now available at git+ssh://git.debian.org/git/pkg-opencl/libclc.git git://anonscm.debian.org/pkg-opencl/libclc.git https://anonscm.debian.org/cgit/pkg-opencl/libclc.git and writable by the members of the pkg-opencl alioth project. Andreas
Bug#792223: mesa-utils: looking for a way to have glxinfo (end friends) installed for both amd64 and i386 at the same time
Package: mesa-utils Version: 8.2.0-1 Severity: wishlist For debugging 32-bit GL issues on a amd64 host it may be useful to have access to glxinfo for amd64 and for i386 at the same time. One possible solution could be to ship the binaries renamed to ${glxbinary}-${DEB_HOST_MULTIARCH} (or ${DEB_HOST_MULTIARCH}-${glxbinary} as the GCC toolchain does), maybe in a new binary package 'mesa-utils-bin' that is Multi-Arch:same, and have mesa-utils ship ${glxbinary} - ${DEB_HOST_MULTIARCH}-${glxbinary} symlinks while depending on mesa-utils-bin:native. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20150712200415.14655.99124.report...@zam581.zam.kfa-juelich.de
Bug#769191: Bug#769072: , #769191, #770588: nvidia-opencl-icd breaking non-nvidia systems
On 2014-11-24 22:47, Rebecca N. Palmer wrote: First of all, many thanks for analyzing the problem and keeping track of the many duplicates! (Should we merge these bugs? Also, #767803 looks like another instance of this, though it doesn't have the apt log to confirm it yet) and perhaps move the discussion to one of these bugs and trim the Cc list a bit ... If you merge them, please reassign them to src:nvidia-driver and add Affects for the packages where they have been reported. * nvidia-kernel-dkms: Switch to Recommends: nvidia-driver | libcuda1 to break the chain libcuda1 - nvidia-kernel-dkms - nvidia-driver. ...or drop this Recommends: entirely That Recommends has been there since nvidia-kernel-dkms was added in 2010. I thinks its main purpose is for people that try to install the wrong package (i.e. nvidia-kernel-dkms instead of nvidia-driver (or nvidia-glx back then)) to get a working NVIDIA driver installation (well, still needs some manual configuration, but debconf will tell you that). I don't want to touch (as in remove) that right now. (IIRC circular Depends/Recommends are discouraged because they confuse apt's autoremover, though I can't find where I saw that). I don't want to workaround bugs in unrelated packages Cutting the chain here (tested with apt-get install nvidia-libopencl1 nvidia-driver-, the - after a package means remove/don't install) does still allow much of nvidia-* (including nvidia-kernel-dkms and glx-alternative-nvidia) to be installed, but that appears to be harmless and you get the full set of lib*GL* diversions, too ... this and some more are needed to allow switching between nvidia-opencl-icd and nvidia$LEGACY-opencl-icd which will become available once I fork off (and backport) nvidia-graphics-drivers-legacy-340xx after the jessie release. without libgl1-nvidia-glx (at least on my Intel IvyBridge M GT2, both graphics and OpenCL continue to work after rebooting). That would be a bug otherwise :-) Given that the error on loading nvidia-opencl-icd in a non-Nvidia system is modprobe: ERROR: could not insert 'nvidia_current': No such device it is plausible that nvidia-opencl-icd uses the nvidia kernel module (i.e. nvidia-kernel-dkms | nvidia-kernel-version) and as such _should_ pull it in (whether or not it also needs libcuda1), In wheezy all this OpenCL stuff was rather new and only proprietary implementations were available ... as a result installing nvidia-opencl-icd was harmless (due to lack of dependencies) but also useless (due to lack of dependencies) unless you had nvidia-driver installed, too. This I wanted to fix in jessie :-) while #768185 suggests that nvidia-opencl-icd works without the graphics side (can someone check that?), I'm pretty sure NVIDIA CUDA and OpenCL work in headless setups, i.e. without X server running. They just need the kernel module to be loaded (and the corresponding device to be created). (Contrary to AMD which needs X to communicate with the GPU for doing OpenCL) making this the more correct place to cut the chain. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/547472c0.3030...@debian.org
Bug#769191: Bug#769072: #769191: nvidia-opencl-icd breaking non-nvidia systems
On 2014-11-23 00:30, Rebecca N. Palmer wrote: I think the trigger is nvidia-opencl-icd adding a new dependency on libcuda1 (changelog: Add libcuda1 dependency to libraries that seem to be capable of doing dlopen(libcuda.so) or dlopen(libcuda.so.1).), which pulls in the rest of nvidia-* as libcuda1 Recommends: nvidia-kernel-dkms which Recommends: nvidia-driver. Damn, ensuring that there are proper dependencies for all the NVIDIA packages breaks (well, sort of breaks) the upgrade in case nvidia-opencl-icd was installed in wheezy for no good reason ... I've now prepared two changes in SVN (n-g-d branches/jessie): nvidia-graphics-drivers (340.46-6) UNRELEASED; urgency=medium * nvidia-kernel-dkms: Switch to Recommends: nvidia-driver | libcuda1 to break the chain libcuda1 - nvidia-kernel-dkms - nvidia-driver. * nvidia-opencl-icd: Downgrade the Depends: libcuda1 to Suggests. This should avoid pulling in too many NVIDIA packages on wheezy - jessie upgrades of systems that have no NVIDIA hardware, but nvidia-opencl-icd installed nevertheless. (Closes: #769072, #770588, #769191) In upgrade tests in minimal wheezy chroots with nvidia-opencl-icd installed that seemed to work and did no longer pull nvidia-driver into the system. I don't know how seriously the missing libcuda1 breaks nvidia-opencl-icd. I can see that this is being dlopen()ed, but at least clinfo still reports something about the GPU. I don't have a better testcase right now, suggestions welcome. A more proper solution could be to rename nvidia-opencl-icd to e.g. nvidia-icd and add an empty nvidia-opencl-icd package that only Suggests: nvidia-icd. That would prevent inadvertent users of nvidia-opencl-icd from getting anything new from the nvidia stack but everyone who really wants to use nvidia-icd would have to install it manually. (but we could Recommend it from nvidia-opencl-dev (src:nvidia-cuda-toolkit)). Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/547141d9.9050...@debian.org
Re: Bug#769072: gnome-shell: GNOME Shell crashing with Oh no, something went wrong
On 2014-11-16 09:13, Janusz S. Bien wrote: Quote/Cytat - Andreas Beckmann a...@debian.org (nie, 16 lis 2014, (I would guess neither nvidia-driver nor xserver-xorg-video-nvidia). I was right :-) Although some of them has been installed manually, for the first time the nvidia packages just appeared after one of (dist?) upgrades. The first trace of them is from August: Mon, Aug 4 2014 16:15:15 +0200 [INSTALL, DEPENDENCIES] nvidia-libopencl1:i386 [INSTALL, DEPENDENCIES] nvidia-opencl-icd:i386 Something OpenCL related pulled these in - I think we have changed the dependencies since then on the OpenCL parts globally to prevent pulling in bits of the non-free drivers just for getting a libOpenCL.so.1 as some dependency (you can manually select to install the proprietary ones, though). Can I remove all of the packages listed above? these and glx-diversions and glx-alternative-*, maybe some more :-) If something needs a libopencl1, use ocl-icd-libopencl1 instead of the nvidia one. I'm still intrigued by the message which I definitely seen on the console during an upgrade, which was definitely about a nvidia package and which is not present in /var/apt/log. The problem appeared immediately after this routine upgrade. Perhaps by my actions I made it worst, but the primary reason was (is?) somewhere in the system. I don't think any of the debconf messages used by the nvidia driver would have been displayed in your case. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/54687871.2080...@debian.org
Re: Bug#769072: gnome-shell: GNOME Shell crashing with Oh no, something went wrong
On 2014-11-15 14:13, Julien Cristau wrote: Looks like nvidia took over libGL but not the server-side libglx.so. That can't possibly work; can the nvidia packages not do that? It should ... (not diverting, but placing its libglx.so in the front of the search path, controlled by alternatives) Please send the output from the bug script from the nvidia packages /usr/share/bug/libgl1-nvidia-glx/script 3nvidia.log On Sat, 15 Nov 2014 08:01:51 +0100 Janusz S. Bien jsb...@mimuw.edu.pl wrote: As I wrote already, there was a message during one of the upgrades which I unfortunately ignored. Should it be still available somewhere in the logs? I tried to find it, but unsuccesfully. Try /var/log/apt/term.log* Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/5467b6dc.9050...@debian.org
Re: Bug#769072: gnome-shell: GNOME Shell crashing with Oh no, something went wrong
On 2014-11-15 21:53, Janusz S. Bien wrote: Quote/Cytat - Andreas Beckmann a...@debian.org (sob, 15 lis 2014, /usr/share/bug/libgl1-nvidia-glx/script 3nvidia.log Enclosed. Thanks. Does not look like something is broken (except for having the nvidia packages installed manually on a system that does not support them). Which nvidia packages do you have installed? dpkg -l | grep nvidia (I would guess neither nvidia-driver nor xserver-xorg-video-nvidia). To fix your system, either uninstall the nvidia packages (they are not useful on your system anyway) or switch back to mesa: update-alternatives --config glx and select mesa I don't think I can express this with package relationships: libgl1-nvidia-glx needs to be installed (1) together with xserver-xorg-video-nvidia (and xserver-xorg-core) (regular installation on a host) (2) OR without xserver-xorg-core (and xserver-xorg-video-nvidia) (i.e. libraries from a foreign architecture only or in a chroot without own xserver) i.e. only the combination install libgl1-nvidia-glx:native install xserver-xorg-core:native no-install xserver-xorg-video-nvidia should be forbidden. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/5467ef4d.2050...@debian.org
Bug#765967: [libvdpau1] mpeg 1 video are decoded too bad
On 2014-10-19 21:36, Vincent Bernat wrote: Nope, its string OpenGL/VAAPI/libswscale backend for VDPAU. I suppose that G3DVL means Gallium 3D Video L(ayer). So, this is more likely mesa-vdpau-drivers. OK, I'll add a bug-script to libvdpau1 to list available drivers ... Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/54441708.5030...@debian.org
Re: [Pkg-fglrx-devel] Heads up: xorg-server 1.15 soon in sid
On 2014-01-28 23:39, Julien Cristau wrote: On the closed drivers front, I assume nvidia's non-legacy driver will be fine, and fglrx will lag by a few weeks/months as usual? I just uploaded fglrx-driver 14.1~beta to sid that adds support for Xserver 1.15, so that one should be OK, too. (The tracker won't reflect this since the non-free drivers depend on an OR-ed list of all supported server versions and therefore match the bad expression, too.) Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/52fcd6be.7020...@debian.org
Re: Heads up: xorg-server 1.15 soon in sid
On 2014-01-28 23:39, Julien Cristau wrote: On the closed drivers front, I assume nvidia's non-legacy driver will be fine, This time nvidia was quick: all (legacy) drivers currently in jessie already support 1.15. and fglrx will lag by a few weeks/months as usual? Probably ... What is the X roadmap for jessie? Will we likely see another update? Xserver 1.16 is planned for July [1] [1] http://lists.x.org/archives/xorg-devel/2013-December/039739.html Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/52e86db8.9010...@debian.org
Re: heads up: xserver 1.14 to unstable
On 2013-09-17 21:27, Julien Cristau wrote: I notice that fglrx-driver, fglrx-legacy-driver and nvidia-graphics-drivers-legacy-96xx in sid don't declare support for the new ABI. Is there a newer version of those? nvidia-graphics-drivers-legacy-96xx probably won't see any updates from upstream, so we may need to phase it out for jessie fglrx-driver in experimental has support for 1.14 fglrx-legacy-driver may or may not get an update from upstream Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/523ac196.3050...@debian.org
Bug#720069: libgl1-mesa-glx: add Breaks: glx-diversions ( 0.4)
Package: libgl1-mesa-glx Version: 9.1.6-3 Severity: normal Tags: patch Hi, please add a Breaks: glx-diversions ( 0.4) to libgl1-mesa-glx to ensure clean upgrades from MESA 8 to MESA 9 if the proprietary drivers are being used. MESA 9 ships libGL.so.1.2.0 which needs to be diverted by glx-diversions, too. Patch will follow. Thanks Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130818084436.4050.54557.report...@cake.ae.cs.uni-frankfurt.de
Bug#720069: [PATCH] libgl1-mesa-glx needs Breaks: glx-diversions ( 0.4)
the old glx-diversions versions are not fully aware of libGL.so.1.2.0 --- debian/changelog |7 +++ debian/control |1 + 2 files changed, 8 insertions(+) diff --git a/debian/changelog b/debian/changelog index f68dc77..9d03373 100644 --- a/debian/changelog +++ b/debian/changelog @@ -1,3 +1,10 @@ +mesa (9.1.6-3) UNRELEASED; urgency=low + + * libgl1-mesa-glx: Add Breaks: glx-diversions ( 0.4) because the old +versions were not aware of libGL.so.1.2.0. (Closes: #720069) + + -- Andreas Beckmann a...@debian.org Sun, 18 Aug 2013 10:50:28 +0200 + mesa (9.1.6-2) unstable; urgency=low * Don't run 'make -j' if DEB_BUILD_OPTIONS doesn't set parallel. Oops. diff --git a/debian/control b/debian/control index c344e0c..4be6115 100644 --- a/debian/control +++ b/debian/control @@ -543,6 +543,7 @@ Provides: libgl1 Breaks: libgl1-nvidia-alternatives (= 275.09.07-1), fglrx-glx ( 1:11-6-1), + glx-diversions ( 0.4), Pre-Depends: ${misc:Pre-Depends} Multi-Arch: same Description: free implementation of the OpenGL API -- GLX runtime -- 1.7.10.4 -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/1376816258-10897-1-git-send-email-a...@debian.org
Bug#720026: libxfont-dev: arch-dependent file in Multi-Arch: same package
Package: libxfont-dev Version: 1:1.4.6-1 Severity: important User: multiarch-de...@lists.alioth.debian.org Usertags: multiarch libxfont-dev is marked as Multi-Arch: same, but the following file is architecture-dependent: /usr/share/doc/libxfont-dev/fontlib.html An example diff between i386 and amd64 is attached. Cheers, Andreasdiff -ur libxfont-dev_1.4.6-1_i386/usr/share/doc/libxfont-dev/fontlib.html libxfont-dev_1.4.6-1_amd64/usr/share/doc/libxfont-dev/fontlib.html --- libxfont-dev_1.4.6-1_i386/usr/share/doc/libxfont-dev/fontlib.html 2013-08-12 18:40:47.0 +0200 +++ libxfont-dev_1.4.6-1_amd64/usr/share/doc/libxfont-dev/fontlib.html 2013-08-12 18:30:24.0 +0200 @@ -292,7 +292,7 @@ } /style/headbodydiv class=articlediv class=titlepagedivdivh2 class=titlea id=fontlib/a The X Font Library - /h2/divdivdiv class=authorgroupdiv class=authorh3 class=authorspan class=firstnameKeith/span span class=surnamePackard/span/h3div class=affiliationspan class=orgnameMIT X Consortiumbr //span/div/divdiv class=authorh3 class=authorspan class=firstnameDavid/span span class=surnameLemke/span/h3div class=affiliationspan class=orgnameNetwork Computing Devicesbr //span/div/div/div/divdivp class=releaseinfoX Version 11, Release 7.6/p/divdivp class=copyrightCopyright © 1993 Network Computing Devices/p/divdivdiv class=legalnoticea id=idp43674508/ap + /h2/divdivdiv class=authorgroupdiv class=authorh3 class=authorspan class=firstnameKeith/span span class=surnamePackard/span/h3div class=affiliationspan class=orgnameMIT X Consortiumbr //span/div/divdiv class=authorh3 class=authorspan class=firstnameDavid/span span class=surnameLemke/span/h3div class=affiliationspan class=orgnameNetwork Computing Devicesbr //span/div/div/div/divdivp class=releaseinfoX Version 11, Release 7.6/p/divdivp class=copyrightCopyright © 1993 Network Computing Devices/p/divdivdiv class=legalnoticea id=idp52162720/ap Permission to use, copy, modify, distribute, and sell this software and its documentation for any purpose is hereby granted without fee, provided that the above copyright
Bug#720026: libxfont-dev: arch-dependent file in Multi-Arch: same package
There is also usr/lib/debug/usr/lib/libXfont.so.1.4.1 in libxfont1-dbg which is different on all architectures. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/520fe23a.10...@debian.org
Bug#706114: xfonts-utils: insufficient zlib1g dependency
On 2013-04-25 10:39, Guillem Jover wrote: On Wed, 2013-04-24 at 22:06:55 -0400, Michael Gilbert wrote: So the dependencies are correct. The only problem is due to gsfonts-x11 postinst's script calling: update-fonts-dir → mkfontdir → mkfontscale While the new xfonts-utils package has been unpacked, but not configured. This is either a problem in xfonts-utils's update-fonts-dir, or in debhelper's dh_installxfonts support inserting those snippets. In terms of getting this fixed quickly for wheezy, couldn't that be addressed with a Pre-Depends: xfonts-utils in gsfonts-x11? That would only solve the problem for gsfonts-x11, but not for anything else using dh_installxfonts. I think, to be safe the hack would need to be applied to xfonts-utils's dependencies, which is all The reason for this suboptimal upgrade order is probably Package: zlib1g Breaks: ..., texlive-binaries ( 2009-12) and instead of deconfiguring texlive-binaries apt delays installation of the new zlib1g, breaking some unpacked and not yet configured packages. kinds of ugly, but at least should always work, at the cost of complicating the upgrade path. I guess the correct solution though, is to change update-fonts-* and any other script in xfonts-utils calling mkfont* from other maintainer scripts, to only run if xfonts-utils itself is configured or being configured, and calling these through xfonts-utils maintainer scripts too, so that if xfonts-utils and something else using it like gsfonts-x11 are both unpacked on the same run, the actual update-fonts-* executions will be delayed until xfonts-utils is itself configured, at which point the scripts should be able to run. I don't think this would be much more difficult to prepare. Sounds like a job for triggers (but there is currently a dpkg bug that causes it to run triggers for packages that are not configured (or their dependencies are not) Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/517926d1.9020...@debian.org
Bug#640499: libxvmc: please add multiarch support
On 2012-12-07 00:43, Carlos Alberto Lopez Perez wrote: MultiArch support is a release goal, and the fix for this bug is clearly not invasive nor big: $ curl -s 'http://bugs.debian.org/cgi-bin/bugreport.cgi?msg=5;filename=0001-build-for-multiarch.patch;att=1;bug=640499' | diffstat 5 files changed, 15 insertions(+), 8 deletions(-) Where is the problem? What prevents the release team allowing this upload for Wheezy? There is a solution that only involves changes to nvidia-graphics-drivers, waiting for approval: splitting libxvmcnvidia1 from libgl1-nvidia-glx: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=688953 new upstream stable branch release http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=688698 Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/50c191e7.1010...@abeckmann.de
Bug#640499: libxvmc: please add multiarch support
On 2012-10-15 21:13, Hussain AlMutawa wrote: Hi, Just reporting. I am using the proprietary nvidia driver on debian sid. Androids Emulator requires the 32 bit version of openGL. Which can not be installed because of this bug. use the nvidia driver packages from unstable which has a workaround Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/507c6ea2.40...@abeckmann.de
Bug#640499: Bug#688861: freeze exception: libxvmc/1.0.7-1.1 - adding a libxvmc1-i386:i386 package
In case anyone is interested in a .dsc: http://mentors.debian.net/debian/pool/main/libx/libxvmc/libxvmc_1.0.7-1.1.dsc Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/5062facc.8070...@abeckmann.de
Bug#640499: Fwd: Bug#688861: freeze exception: libxvmc/1.0.7-1.1 - adding a libxvmc1-i386:i386 package
manually forwarding to #640499 as the BTS does not seem to handle X-DebBugs-Cc: n@b.d.o properly Original Message Subject: Bug#688861: freeze exception: libxvmc/1.0.7-1.1 - adding a libxvmc1-i386:i386 package Resent-Date: Wed, 26 Sep 2012 12:09:06 + Resent-From: Andreas Beckmann deb...@abeckmann.de Resent-To: debian-bugs-d...@lists.debian.org Resent-CC: 640...@bugs.debian.org, pkg-nvidia-de...@lists.alioth.debian.org, Debian Release Team debian-rele...@lists.debian.org Date: Wed, 26 Sep 2012 14:06:33 +0200 From: Andreas Beckmann deb...@abeckmann.de Reply-To: Andreas Beckmann deb...@abeckmann.de, 688...@bugs.debian.org To: Debian Bug Tracking System sub...@bugs.debian.org Package: release.debian.org Severity: normal User: release.debian@packages.debian.org Usertags: unblock Please approve the following changes for package libxvmc As we are not going to get libxvmc turned to multi-arch for wheezy (see #640499) I'm now trying another approach with minimal changes to the libxvmc package: Can we add a new package Package: libxvmc1-i386 Section: libs Priority: extra Architecture: i386 Multi-arch: same Pre-Depends: ${misc:Pre-Depends} Depends: ${shlibs:Depends}, ${misc:Depends}, x11-common Description: X11 Video extension library (alternate i386 package) that ships another copy of the libraries in /usr/lib/i386-linux-gnu/ To avoid installing libxvmc1-i386:i386 and libxvmc1:i386 at the same time and have two copies of the same library available, I'm adding Package: libxvmc1 Conflicts: libxvmc1-i386 [i386] and to avoid picking up a dependency on libxvmc1-i386 by accident, its shlibs file points to libxvmc1. I added a lintian-overrides file, but that is not being installed as I wanted to avoid changing anything in debian/rules. Note that this new package does not ship the conffile as it is expected to be only installed along libxvmc1:amd64 which comes with the conffile. (This dependency can't be expressed by package relationships.) This will keep the existing non-multiarch libxvmc1 unchanged, the build system is untouched, nothing will start to use the alternate copy package accidently. For libgl1-nvidia-glx:i386 to actually pick up this alternate library, we need to add a shlibs.local file (in nvidia-graphics-drivers) with libXvMC 1 libxvmc1 | libxvmc1-i386 [i386] I chose to add the extra package to the libxvmc source package as that will avoid forking (and therefore code duplication) and getting libxvmc1 and libxvmc1-i386 out of sync. Even with a second copy in the a binary package in the archive, there is no overhead in case a security update for libxvmc should be neccessary. Judging by the number of bug reports, it is really important for our users to restore the ability to run 32-bit OpenGL applications on a amd64 system and get accelleration by the non-free nvidia driver. That was working in squeeze and will be considered as a serious regression if this is not getting fixed in some way for wheezy. See e.g. #685054, #686033, #676723, #688714 I'll be happy to assist getting a proper multi-arch libxvmc into jessie and to clean up this temporary package that we need for wheezy. Andreas diffstat for libxvmc_1.0.7-1 libxvmc_1.0.7-1.1 debian/libxvmc1-i386.install |2 ++ debian/libxvmc1-i386.lintian-overrides |4 debian/libxvmc1-i386.shlibs|2 ++ libxvmc-1.0.7/debian/changelog | 12 libxvmc-1.0.7/debian/control | 26 ++ 5 files changed, 46 insertions(+) diff -u libxvmc-1.0.7/debian/control libxvmc-1.0.7/debian/control --- libxvmc-1.0.7/debian/control +++ libxvmc-1.0.7/debian/control @@ -22,6 +22,7 @@ Section: libs Architecture: any Depends: ${shlibs:Depends}, ${misc:Depends}, x11-common +Conflicts: libxvmc1-i386 [i386] Description: X11 Video extension library libXvMC provides an X Window System client interface to the XVideo-MotionCompensation extension to the X protocol. @@ -37,6 +38,31 @@ This module can be found at git://anongit.freedesktop.org/git/xorg/lib/libXvMC +Package: libxvmc1-i386 +Section: libs +Priority: extra +Architecture: i386 +Multi-arch: same +Pre-Depends: ${misc:Pre-Depends} +Depends: ${shlibs:Depends}, ${misc:Depends}, x11-common +Description: X11 Video extension library (alternate i386 package) + libXvMC provides an X Window System client interface to the + XVideo-MotionCompensation extension to the X protocol. + . + The XVideo-MotionCompensation extension allows for further accelerated drawing + of videos. Video data may be sent at earlier stages of the decoding pipeline + than raw YUV data. At the moment, driver support for XvMC is poor to + non-existent. + . + More information about X.Org can be found at: + URL:http://www.X.org + . + This module can be found at + git://anongit.freedesktop.org/git/xorg/lib/libXvMC
Bug#640499: libxvmc: please add multiarch support
On Tuesday, 14. August 2012 11:35:21 Ralf Jung wrote: As it turned out, the conffile issue is actually a bug in dpkg [1]. In theory (and once dpkg is fixed), it is perfectly fine to have a conffile in an MA: same package, as long as the content is the same for all architectures. Have the upgrade and install paths paths of this scenario been tested? And the case of locally modified/removed config files? I'll take libfoo as an example package that ships /etc/foo.conf apt-get install libfoo:arch1 modify-or-delete /etc/foo.conf apt-get install libfoo:arch2 # libfoo=1 is available apt-get install libfoo:arch1=1 libfoo:arch2=1 modify-or-delete /etc/foo.conf apt-get update # makes libfoo=2 available # libfoo=2 comes with a modified foo.conf apt-get dist-upgrade # libfoo=1 is available apt-get install libfoo:arch1=1 libfoo:arch2=1 apt-get update # makes libfoo=2 available # libfoo=2 uses dpkg-maintscript-helper to get rid of foo.conf apt-get dist-upgrade Just curious ... Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/201208180109.11567.deb...@abeckmann.de
Bug#682939: xfonts-utils, xfonts-encodings, debhelper: installing xfonts-tipa removes /usr/share/fonts/X11/encodings/encodings.dir
Package: xfonts-utils,xfonts-encodings,debhelper Severity: important User: debian...@lists.debian.org Usertags: piuparts Hi, during a test with piuparts I noticed the xfonts-tipa packages removes /usr/share/fonts/X11/encodings/encodings.dir which is owned by xfonts-encodings. xfonts-tipa maintainer scripts are entirely debhelper generated, so the xfonts-tipa package is not doing something wrong, this is part of the postinst: # Automatically added by dh_installxfonts if which update-fonts-dir /dev/null 21; then update-fonts-dir --x11r7-layout encodings;update-fonts-scale Type1;update-fonts-dir --x11r7-layout Type1 fi # End automatically added section Thereafter: debsums: missing file /usr/share/fonts/X11/encodings/encodings.dir (from xfonts-encodings package) Is it correct that encodings.dir is shipped by xfonts-encodings? Shouldn't that be generated by maintainer scripts? Otherwise the update-fonts-dir --x11r7-layout encodings will always modify shipped files. Why does the encodings.dir file get deleted at all? That directory is not empty ... I'm assigning this bug to three packages that could possibly be at fault: * xfonts-utils - update-fonts-dir deletes encodings.dir, but dir is not empty * xfonts-encodings - ships encodings.dir * debhelper - could have generated a wrong call Please analyze and reassign. piuparts log from xfonts-tipa is attached. Andreas xfonts-tipa_2:1.3-17.log.gz Description: GNU Zip compressed data
Bug#682939: xfonts-utils, xfonts-encodings, debhelper: installing xfonts-tipa removes /usr/share/fonts/X11/encodings/encodings.dir
On 2012-07-27 13:22, Daniele Tricoli wrote: piuparts seems to be ok now (logfile attached). You can't rerun the test with the released piuparts, only with the version from git. But it's easy to test without piuparts: install the package in a clean minimal chroot (e.g. pbuilder, cowbuilder) /usr/share/fonts/X11/encodings/encodings.dir should not disappear, test with debsums xfonts-encodings Andreas PS: You probably want to do this, too: bts reassign 682939 xfonts-tipa 2:1.3-17 -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/50127cc9.2010...@abeckmann.de
Bug#674563: mesa: install libGL.so* into $libdir/mesa, add links to $libdir
Source: mesa Version: 7.11.2-1 Severity: wishlist Tags: patch Hi, this is in response to a thread started here: http://lists.debian.org/debian-x/2011/07/msg00292.html and later continued here: http://lists.debian.org/debian-x/2011/10/msg00137.html In order to simplify managing alternative GL implementations, I propose to ship the MESA libraries in $libdir/mesa/ with symlinks to libGL.so.1 and libGL.so in $libdir/ This is a first start and helps to reduce the number of diversions needed. The next step would be the use of some alternatives system directly in the MESA packages (that needs to be multi-arch aware). This will remove the need for diversions at all, but needs more discussion ... I also sent some patches to debian-x@ but never submitted a bug report, so these patches for moving libGL* probably were forgotten: http://lists.debian.org/debian-x/2011/10/msg00149.html (for GL) http://lists.debian.org/debian-x/2011/10/msg00378.html (for EGL, GLES) If anyone is interested, I'll update the patches for MESA 8. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120525125648.32290.48480.report...@cake.ae.cs.uni-frankfurt.de
video drivers in experimental that depend on outdated video-abi 11
Hi, there are two xorg video drivers in experimental that were built against video-abi versions older than 11: xserver-xorg-video-v4l -- xorg-video-abi-8.0 this package was removed from unstable in #612191, but the copy in experimental was probably missed xserver-xorg-video-r128 -- xorg-video-abi-10 the same version is in sid (except for the +exp? suffix) but had some binNMUs inbetween Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4f5d4fd3.8040...@abeckmann.de
Re: New X stack backport?
On 2012-01-12 12:27, Cyril Brulebois wrote: I'm still not too sure whether to replace the backported X stack entirely or whether to publish a squeeze-backports repository on my own website with source + binaries for i386 and amd64, so that people can try 1.10 from squeeze-backports, or 1.11 from my repository if that's not enough. Maintaining the non-free graphics drivers (nvidia, nvidia-legacy-173xx, nvidia-legacy-96xx, fglrx), I'd like to keep the old Xserver versions available. The proprietary drivers are compatible with all old video-abi versions, but it usually takes some time until they add support for new Xserver releases (which bump the abi). Right now Xserver 1.10 from squeeze-backports is needed for the nvidia-legacy drivers in sid (upstream has not yet abandoned them, but not yet made a release supporting 1.11 in wheezy/sid) and I'm considering backporting the legacy drivers to squeeze-backports, too (would e.g. add support for the backports kernel and we could point users to a place where a working combination of Xorg + legacy driver is available). Current nvidia-glx in squeeze-backports, wheezy and sid work in 1.11, but there may be issues that do not occur in 1.10. fglrx now supports 1.11, but there still seem to be issues ... (fglrx backport is ancient ... requiring Xserver from squeeze not backports) Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4f0ef922.5060...@abeckmann.de
[PATCH] install libEGL* and libGLES* into $libdir/mesa
Install libEGL.so*, libGLESv1_CM.so*, and libGLESv2.so* into $libdir/mesa/ instead of $libdir/ and add symlinks for lib*.so and lib*.so.[12] (but not lib*.so.[12].*) to $libdir. This is a preparation for supporting alternative implementations of libEGL and libGLES (currently only a non-free alternative from NVIDIA exists). The links in $libdir will be managed by update-alternatives later on and therefore it is necessary to move the real files (lib*.so.[12].*) out of $libdir, otherwise ldconfig will happily modify the lib*.so.[12] SONAME links. Signed-off-by: Andreas Beckmann deb...@abeckmann.de --- debian/libegl1-mesa-dev.install.in |2 +- debian/libegl1-mesa-dev.links.in|1 + debian/libegl1-mesa.install.in |2 +- debian/libegl1-mesa.links.in|1 + debian/libgles1-mesa-dev.install.in |2 +- debian/libgles1-mesa-dev.links.in |1 + debian/libgles1-mesa.install.in |2 +- debian/libgles1-mesa.links.in |1 + debian/libgles2-mesa-dev.install.in |2 +- debian/libgles2-mesa-dev.links.in |1 + debian/libgles2-mesa.install.in |2 +- debian/libgles2-mesa.links.in |1 + 12 files changed, 12 insertions(+), 6 deletions(-) diff --git a/debian/libegl1-mesa-dev.install.in b/debian/libegl1-mesa-dev.install.in index 3f3094f..5a41338 100644 --- a/debian/libegl1-mesa-dev.install.in +++ b/debian/libegl1-mesa-dev.install.in @@ -1,4 +1,4 @@ -dri/usr/lib/${DEB_HOST_MULTIARCH}/libEGL.so usr/lib/${DEB_HOST_MULTIARCH} +dri/usr/lib/${DEB_HOST_MULTIARCH}/libEGL.so usr/lib/${DEB_HOST_MULTIARCH}/mesa #dri/usr/lib/${DEB_HOST_MULTIARCH}/libwayland-egl.so usr/lib/${DEB_HOST_MULTIARCH} dri/usr/include/EGL usr/include dri/usr/include/KHR usr/include diff --git a/debian/libegl1-mesa-dev.links.in b/debian/libegl1-mesa-dev.links.in new file mode 100644 index 000..da31542 --- /dev/null +++ b/debian/libegl1-mesa-dev.links.in @@ -0,0 +1 @@ +usr/lib/${DEB_HOST_MULTIARCH}/mesa/libEGL.so usr/lib/${DEB_HOST_MULTIARCH}/libEGL.so diff --git a/debian/libegl1-mesa.install.in b/debian/libegl1-mesa.install.in index 57d1a21..f6ef698 100644 --- a/debian/libegl1-mesa.install.in +++ b/debian/libegl1-mesa.install.in @@ -1 +1 @@ -dri/usr/lib/${DEB_HOST_MULTIARCH}/libEGL.so.1* usr/lib/${DEB_HOST_MULTIARCH} +dri/usr/lib/${DEB_HOST_MULTIARCH}/libEGL.so.1* usr/lib/${DEB_HOST_MULTIARCH}/mesa diff --git a/debian/libegl1-mesa.links.in b/debian/libegl1-mesa.links.in new file mode 100644 index 000..537e2c6 --- /dev/null +++ b/debian/libegl1-mesa.links.in @@ -0,0 +1 @@ +usr/lib/${DEB_HOST_MULTIARCH}/mesa/libEGL.so.1 usr/lib/${DEB_HOST_MULTIARCH}/libEGL.so.1 diff --git a/debian/libgles1-mesa-dev.install.in b/debian/libgles1-mesa-dev.install.in index 0485b23..df6458d 100644 --- a/debian/libgles1-mesa-dev.install.in +++ b/debian/libgles1-mesa-dev.install.in @@ -1,3 +1,3 @@ -dri/usr/lib/${DEB_HOST_MULTIARCH}/libGLESv1_CM.so usr/lib/${DEB_HOST_MULTIARCH} +dri/usr/lib/${DEB_HOST_MULTIARCH}/libGLESv1_CM.so usr/lib/${DEB_HOST_MULTIARCH}/mesa dri/usr/include/GLES usr/include dri/usr/lib/${DEB_HOST_MULTIARCH}/pkgconfig/glesv1_cm.pc usr/lib/${DEB_HOST_MULTIARCH}/pkgconfig diff --git a/debian/libgles1-mesa-dev.links.in b/debian/libgles1-mesa-dev.links.in new file mode 100644 index 000..3d5288b --- /dev/null +++ b/debian/libgles1-mesa-dev.links.in @@ -0,0 +1 @@ +usr/lib/${DEB_HOST_MULTIARCH}/mesa/libGLESv1_CM.so usr/lib/${DEB_HOST_MULTIARCH}/libGLESv1_CM.so diff --git a/debian/libgles1-mesa.install.in b/debian/libgles1-mesa.install.in index 143abd2..aa568d8 100644 --- a/debian/libgles1-mesa.install.in +++ b/debian/libgles1-mesa.install.in @@ -1 +1 @@ -dri/usr/lib/${DEB_HOST_MULTIARCH}/libGLESv1_CM.so.1* usr/lib/${DEB_HOST_MULTIARCH} +dri/usr/lib/${DEB_HOST_MULTIARCH}/libGLESv1_CM.so.1* usr/lib/${DEB_HOST_MULTIARCH}/mesa diff --git a/debian/libgles1-mesa.links.in b/debian/libgles1-mesa.links.in new file mode 100644 index 000..f6b8d8a --- /dev/null +++ b/debian/libgles1-mesa.links.in @@ -0,0 +1 @@ +usr/lib/${DEB_HOST_MULTIARCH}/mesa/libGLESv1_CM.so.1 usr/lib/${DEB_HOST_MULTIARCH}/libGLESv1_CM.so.1 diff --git a/debian/libgles2-mesa-dev.install.in b/debian/libgles2-mesa-dev.install.in index ae8fe70..eaae986 100644 --- a/debian/libgles2-mesa-dev.install.in +++ b/debian/libgles2-mesa-dev.install.in @@ -1,3 +1,3 @@ -dri/usr/lib/${DEB_HOST_MULTIARCH}/libGLESv2.so usr/lib/${DEB_HOST_MULTIARCH} +dri/usr/lib/${DEB_HOST_MULTIARCH}/libGLESv2.so usr/lib/${DEB_HOST_MULTIARCH}/mesa dri/usr/include/GLES2 usr/include dri/usr/lib/${DEB_HOST_MULTIARCH}/pkgconfig/glesv2.pc usr/lib/${DEB_HOST_MULTIARCH}/pkgconfig diff --git a/debian/libgles2-mesa-dev.links.in b/debian/libgles2-mesa-dev.links.in new file mode 100644 index 000..328ade9 --- /dev/null +++ b/debian/libgles2-mesa-dev.links.in @@ -0,0 +1 @@ +usr/lib/${DEB_HOST_MULTIARCH}/mesa/libGLESv2.so usr/lib/${DEB_HOST_MULTIARCH}/libGLESv2.so diff --git a/debian/libgles2-mesa.install.in b/debian/libgles2-mesa.install.in
README.source and documentation updates (was: Re: mesa: Changes to 'debian-unstable')
commit aee32cbeae3e76bab261fc02fa5e608b71360439 Author: Cyril Brulebois k...@debian.org Date: Sat Sep 24 20:32:27 2011 +0200 Document the symlink dance in README.source. Thanks! This was giving me a headache previously. Eventually this can be extended by an example how to set up .git/gbp.conf so that it does export the git tree and run the symlink dance automatically? This works for me: [git-buildpackage] prebuild = find -type l | while read dest; do src=$(readlink -f $dest); rm $dest; cp $src $dest; done export-dir = ../build-area/ tarball-dir = ../tarballs/ Also it would be nice if README.source would have a link to http://pkg-xorg.alioth.debian.org/ (and eventually other places with XSF packaging related information) - this probably applies to all packages managed by XSF. Makes it easier for non-XSF people to start hacking X properly :-) Something I didn't find so far are instructions how you would like to get patches submitted. Some minor updates for http://wiki.debian.org/Xorg: * wheezy has xserver-xorg-core 1.11 * In the Video Drivers section, can you add a link to http://wiki.debian.org/NvidiaGraphicsDrivers? (This page mainly talks about the non-free driver). Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4e9c00c6.6030...@abeckmann.de
Re: [RFC] proposal for reorganizing the MESA libraries to simplify replacing them with vendor implementations
Hi all, let's pick up this discussion again. On 2011-07-25 19:03, Julien Cristau wrote: On Mon, Jul 25, 2011 at 18:57:17 +0200, Andreas Beckmann wrote: What is the correct approach to do alternatives of libraries in Multi-Arch: yes packages? * a separate alternative for each architecture (allows weird mixed configurations) This is probably the way to go since there may be many valid combinations thanks to multiarch. Eventually a new update-glx utility could be introduced that allows easy configuration of sane combinations: update-glx --set nvidia which would update the libGL.so.1 alternatives for i386 and amd64 (if available) * one alternative (from an extra Multi-Arch: foreign package) that covers all arches as slaves links? That's the way currently done by glx-alternatives - but it only covers amd64 and i386, a generic extension will be difficult. * the Ubuntu approach of dropping something in ld.so.conf.d/ - would we ... Pretty please not messing with ld.so.conf. OK, lets forget about this. I'm about to send a patch for MESA which moves libGL.so* from $libdir to $libdir/mesa/ and adds libGL.so and libGL.so.1 symlinks from $libdir to the file in $libdir/mesa/. This is a preparative step, the links in $libdir are to be replaced by alternatives later on. For now, they will still be diverted by glx-diversions. glx-alternative-mesa has already been updated to take $libdir/mesa into account. To effectively use diversions, the real files (libGL.so.1.*) need to be moved out of $libdir, otherwise ldconfig may mess up the SONAME links. With the current change it will already be possible to stop diverting libGL.so. I did not yet look into EGL, but I think it needs to be done similarily there - move all libs that may get alternatives to $libdir/mesa/ Does XSF plan to backport MESA 7.11 to squeeze-backports? Is there a planned timeframe? If my patch for the directory structure is accepted and will go to squeeze-backports, too, it would significantly simplify my plans for backporting nvidia-graphics-drivers. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4e9c1a11.7030...@abeckmann.de
[PATCH] install libGL.so* into $libdir/mesa, add links to $libdir
Install libGL.so* into $libdir/mesa/ instead of $libdir/ and add symlinks for libGL.so and libGL.so.1 (but not libGL.so.1.*) to $libdir. This is a preparation for simplifying the usage (aka avoid diversions) of alternative implementations of libGL (currently only non-free alternatives from NVIDIA and AMD exist). The links in $libdir will be managed by update-alternatives later on and therefore it is necessary to move the real file (libGL.so.1.*) out of $libdir, otherwise ldconfig will happily modify the libGL.so.1 SONAME link. Signed-off-by: Andreas Beckmann deb...@abeckmann.de --- debian/libgl1-mesa-dev.install.in |2 +- debian/libgl1-mesa-dev.links.in |1 + debian/libgl1-mesa-glx.install.in |2 +- debian/libgl1-mesa-glx.links.in |1 + debian/libgl1-mesa-swx11-dev.install.in |2 +- debian/libgl1-mesa-swx11-dev.links.in |1 + debian/libgl1-mesa-swx11.install.in |2 +- debian/libgl1-mesa-swx11.links.in |1 + 8 files changed, 8 insertions(+), 4 deletions(-) diff --git a/debian/libgl1-mesa-dev.install.in b/debian/libgl1-mesa-dev.install.in index d915191..5521ac9 100644 --- a/debian/libgl1-mesa-dev.install.in +++ b/debian/libgl1-mesa-dev.install.in @@ -1,2 +1,2 @@ -usr/lib/${DEB_HOST_MULTIARCH}/libGL.so +usr/lib/${DEB_HOST_MULTIARCH}/libGL.so usr/lib/${DEB_HOST_MULTIARCH}/mesa/ usr/lib/${DEB_HOST_MULTIARCH}/pkgconfig/gl.pc diff --git a/debian/libgl1-mesa-dev.links.in b/debian/libgl1-mesa-dev.links.in new file mode 100644 index 000..3bedd3b --- /dev/null +++ b/debian/libgl1-mesa-dev.links.in @@ -0,0 +1 @@ +usr/lib/${DEB_HOST_MULTIARCH}/mesa/libGL.so usr/lib/${DEB_HOST_MULTIARCH}/libGL.so diff --git a/debian/libgl1-mesa-glx.install.in b/debian/libgl1-mesa-glx.install.in index f5ffd7a..2f5e1ad 100644 --- a/debian/libgl1-mesa-glx.install.in +++ b/debian/libgl1-mesa-glx.install.in @@ -1 +1 @@ -dri/usr/lib/${DEB_HOST_MULTIARCH}/libGL.so.* usr/lib/${DEB_HOST_MULTIARCH} +dri/usr/lib/${DEB_HOST_MULTIARCH}/libGL.so.* usr/lib/${DEB_HOST_MULTIARCH}/mesa/ diff --git a/debian/libgl1-mesa-glx.links.in b/debian/libgl1-mesa-glx.links.in new file mode 100644 index 000..4cc810f --- /dev/null +++ b/debian/libgl1-mesa-glx.links.in @@ -0,0 +1 @@ +usr/lib/${DEB_HOST_MULTIARCH}/mesa/libGL.so.1 usr/lib/${DEB_HOST_MULTIARCH}/libGL.so.1 diff --git a/debian/libgl1-mesa-swx11-dev.install.in b/debian/libgl1-mesa-swx11-dev.install.in index 8318ac8..73aef2a 100644 --- a/debian/libgl1-mesa-swx11-dev.install.in +++ b/debian/libgl1-mesa-swx11-dev.install.in @@ -1,2 +1,2 @@ usr/lib/${DEB_HOST_MULTIARCH}/libGL.a -usr/lib/${DEB_HOST_MULTIARCH}/libGL.so +usr/lib/${DEB_HOST_MULTIARCH}/libGL.so usr/lib/${DEB_HOST_MULTIARCH}/mesa/ diff --git a/debian/libgl1-mesa-swx11-dev.links.in b/debian/libgl1-mesa-swx11-dev.links.in new file mode 100644 index 000..3bedd3b --- /dev/null +++ b/debian/libgl1-mesa-swx11-dev.links.in @@ -0,0 +1 @@ +usr/lib/${DEB_HOST_MULTIARCH}/mesa/libGL.so usr/lib/${DEB_HOST_MULTIARCH}/libGL.so diff --git a/debian/libgl1-mesa-swx11.install.in b/debian/libgl1-mesa-swx11.install.in index d333e7b..9d89c20 100644 --- a/debian/libgl1-mesa-swx11.install.in +++ b/debian/libgl1-mesa-swx11.install.in @@ -1 +1 @@ -usr/lib/${DEB_HOST_MULTIARCH}/libGL.so.* +usr/lib/${DEB_HOST_MULTIARCH}/libGL.so.* usr/lib/${DEB_HOST_MULTIARCH}/mesa/ diff --git a/debian/libgl1-mesa-swx11.links.in b/debian/libgl1-mesa-swx11.links.in new file mode 100644 index 000..4cc810f --- /dev/null +++ b/debian/libgl1-mesa-swx11.links.in @@ -0,0 +1 @@ +usr/lib/${DEB_HOST_MULTIARCH}/mesa/libGL.so.1 usr/lib/${DEB_HOST_MULTIARCH}/libGL.so.1 -- tg: (63e119d..) t/use-libdir-mesa (depends on: debian-unstable) -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/1318855656-23366-1-git-send-email-deb...@abeckmann.de
Bug#641344: rendering errors and crashes with nvidia driver and xserver 1.11 due to wrong symbols in libwfb.so
On 2011-09-20 14:20, Cyril Brulebois wrote: Thanks for the tests, folks. We're waiting for the patch to get merged upstream before uploading a fixed package. No need to report more xserver 1.11.1 is out: http://cgit.freedesktop.org/xorg/xserver/log/?h=server-1.11-branch Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4e7d89b9.2010...@abeckmann.de
Bug#641344: Bug#641413: X crashes
On 2011-09-16 13:47, Martin Stigge wrote: Hm, as Tony in #641344 points out, the desktop reacts quite slowly in certain situations with 1.11.0, e.g., switching tabs in pidgin or similar. Downgraded X to 1.10.4 again where this is not the case. That is an independent issue and should be handled in a separate bug report, but Nvidia already seems to investigate this: http://www.nvnews.net/vbulletin/showthread.php?p=2479906#post2479906 Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4e734235.10...@abeckmann.de
Bug#641344: rendering errors and crashes with nvidia driver and xserver 1.11 due to wrong symbols in libwfb.so
subscribe 641344 retitle 641344 rendering errors and crashes with nvidia driver and xserver 1.11 due to wrong symbols in libwfb.so reassign 641413 xserver-xorg-core unblock 641413 by 641344 forcemerge 641344 641413 severity 641344 grave thanks I verified that the attached patch works with xorg-xserver from git and the offending symbols are renamed, but I couldn't verify that it fixes the rendering errors as I cannot reproduce them. I just got confirmation from three users (#641344, #641413) that the patched libwfb.so fixes the issue. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4e73442d.2080...@abeckmann.de
Re: Bug#641344: [nvidia-glx] Graphics errors after upgrading to xserver with abi 11
reassign 641344 xserver-xorg-core 2:1.11.0-1 tag 641344 - moreinfo + patch affects 641344 src:nvidia-graphics-drivers xserver-xorg-video-nvidia retitle 641344 rendering errors with nvidia driver and xserver 1.11 due to wrong symbols in libwfb.so forwarded 641344 http://lists.x.org/archives/xorg-devel/2011-September/025062.html thanks On 2011-09-12 20:25, Franz Schrober wrote: I have a lot of graphics errors after updating my system to the xserver with abi 11. It is hard to write the bug report because I cannot see many gui elements in kde 4. They look like the screen errors I got after using the nvidia driver 285 with the ignore abi option and the abi 11 xorg. NVIDIA tracked this problem down to a bug in libwfb.so and provided a patch: http://www.nvnews.net/vbulletin/showthread.php?t=166025page=2 http://lists.x.org/archives/xorg-devel/2011-September/025062.html http://lists.x.org/archives/xorg-devel/2011-September/025082.html I verified that the attached patch works with xorg-xserver from git and the offending symbols are renamed, but I couldn't verify that it fixes the rendering errors as I cannot reproduce them. Andreas 16-add-missing-libwfb-renames.diff Description: application/pgp-keys
Bug#640499: libxvmc: please add multiarch support
Package: libxvmc Version: 2:1.0.6-1 Severity: wishlist Tags: patch Please rebuild libxvmc with multiarch support. Attached patch is modeled after the multiarch changes in libxv. I'm not sure how to handle /etc/X11/XvMCConfig correctly, might have to be moved to a separate package libxvmc-config? Andreas -- System Information: Debian Release: 6.0.2 APT prefers stable APT policy: (800, 'stable'), (700, 'testing'), (600, 'unstable'), (500, 'stable-updates'), (130, 'experimental') Architecture: amd64 (x86_64) Kernel: Linux 2.6.32-5-amd64 (SMP w/4 CPU cores) Locale: LANG=C, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash From 072c13c64562eafeff15b2c684fa056c3353d592 Mon Sep 17 00:00:00 2001 From: Andreas Beckmann deb...@abeckmann.de Date: Mon, 5 Sep 2011 12:28:14 +0200 Subject: [PATCH] build for multiarch --- debian/changelog |1 + debian/control |5 - debian/libxvmc-dev.install | 10 +- debian/libxvmc1.install|4 ++-- debian/rules |3 +++ 5 files changed, 15 insertions(+), 8 deletions(-) diff --git a/debian/changelog b/debian/changelog index 844e6e7..7b9e6c1 100644 --- a/debian/changelog +++ b/debian/changelog @@ -2,6 +2,7 @@ libxvmc (2:1.0.6-2) UNRELEASED; urgency=low * Rename the build directory to not include DEB_BUILD_GNU_TYPE for no good reason. Thanks, Colin Watson! + * Build for multiarch. -- Cyril Brulebois k...@debian.org Mon, 04 Apr 2011 05:56:30 +0200 diff --git a/debian/control b/debian/control index 7a1914d..9df60c2 100644 --- a/debian/control +++ b/debian/control @@ -4,7 +4,7 @@ Priority: optional Maintainer: Debian X Strike Force debian-x@lists.debian.org Uploaders: David Nusinow dnusi...@debian.org, Andres Salomon dilin...@debian.org, Drew Parsons dpars...@debian.org, Cyril Brulebois k...@debian.org Build-Depends: - debhelper (= 5.0.0), + debhelper (= 8.1.3), libx11-dev (= 1:0.99.2), libxext-dev (= 1:0.99.1), x11proto-video-dev, @@ -21,6 +21,8 @@ Vcs-Browser: http://git.debian.org/?p=pkg-xorg/lib/libxvmc.git Package: libxvmc1 Section: libs Architecture: any +Multi-Arch: same +Pre-Depends: ${misc:Pre-Depends} Depends: ${shlibs:Depends}, ${misc:Depends}, x11-common Description: X11 Video extension library libXvMC provides an X Window System client interface to the @@ -63,6 +65,7 @@ Description: X11 Video extension library (debug package) Package: libxvmc-dev Section: libdevel Architecture: any +Multi-Arch: same Depends: ${shlibs:Depends}, ${misc:Depends}, libxvmc1 (= ${binary:Version}), libx11-dev (= 2:1.0.0-1), libxext-dev (= 1:1.0.0-2), x11proto-video-dev, libxv-dev Conflicts: x11proto-video-dev ( 2.2+cvs.20050712-1) Replaces: x11proto-video-dev ( 2.2+cvs.20050712-1) diff --git a/debian/libxvmc-dev.install b/debian/libxvmc-dev.install index a8b1a8f..ce042b1 100644 --- a/debian/libxvmc-dev.install +++ b/debian/libxvmc-dev.install @@ -1,7 +1,7 @@ usr/include/X11/* -usr/lib/libXvMC.a -usr/lib/libXvMC.so -usr/lib/libXvMCW.a -usr/lib/libXvMCW.so -usr/lib/pkgconfig/xvmc.pc +usr/lib/*/libXvMC.a +usr/lib/*/libXvMC.so +usr/lib/*/libXvMCW.a +usr/lib/*/libXvMCW.so +usr/lib/*/pkgconfig/xvmc.pc usr/share/doc/libXvMC/* usr/share/doc/libxvmc-dev diff --git a/debian/libxvmc1.install b/debian/libxvmc1.install index fe81455..71856ed 100644 --- a/debian/libxvmc1.install +++ b/debian/libxvmc1.install @@ -1,3 +1,3 @@ -usr/lib/libXvMC.so.1* -usr/lib/libXvMCW.so.1* +usr/lib/*/libXvMC.so.1* +usr/lib/*/libXvMCW.so.1* etc/X11/XvMCConfig diff --git a/debian/rules b/debian/rules index 0b5ec6f..17b82d0 100755 --- a/debian/rules +++ b/debian/rules @@ -10,6 +10,8 @@ # set this to the name of the main shlib's binary package PACKAGE = libxvmc1 +DEB_HOST_MULTIARCH ?= $(shell dpkg-architecture -qDEB_HOST_MULTIARCH) + include debian/xsfbs/xsfbs.mk CFLAGS = -Wall -g @@ -40,6 +42,7 @@ build-stamp: mkdir -p build cd build \ ../configure --prefix=/usr \ +--libdir=\$${prefix}/lib/$(DEB_HOST_MULTIARCH) \ --sysconfdir=/etc --mandir=\$${prefix}/share/man \ --infodir=\$${prefix}/share/info $(confflags) \ CFLAGS=$(CFLAGS) -- 1.7.2.5
Re: Xorg module searchpath override directory
Hi all, since I didn't get any feedback, yet, I'm asking again. Are there any objections to installing the alternative slave link for vendor implemntations of libglx.so as /usr/lib/xorg/modules/linux/libglx.so and leave /usr/lib/xorg/modules/extensions/libglx.so undiverted? According to the source code and my tests, the linux subdirectory is always searched first for modules and therefore suitable to install override modules. Andreas On 2011-08-10 23:08, Andreas Beckmann wrote: On 2011-07-25 19:03, Julien Cristau wrote: On Mon, Jul 25, 2011 at 18:57:17 +0200, Andreas Beckmann wrote: libGL.so.1 is not the only file to be diverted, libglx.so is being replaced by the vendors drivers, too. Or is there some way of search path magic that would allow to place the higher priority alternative in a different location where it would be picked up before /usr/lib/xorg/modules/extensions/libglx.so ? That would allow to avoid the diversion here. Something similar to /lib/modules/*/updates/ ... That shouldn't be too hard. I tried to look in the code (xorg-server (2:1.7.7-13) hw/xfree86/loader/* ; the experimental version does not differ much in that area) but didn't quite get how the search path is working. Therefore I added a bit debugging code to trace FindModuleInSubdir(). Translating to pseudocode this now looks like modulepath=/usr/lib/xorg/modules for subdir in input drivers multimedia extensions internal do search_recursive ${modulepath}/${subdir}/linux/ search_recursive ${modulepath}/${subdir}/ done So the order the directories are traversed and files are found depends on the raw order in the filesystem (opendir()), except for the linux subdir which is checked first. hw/xfree86/loader/loadmod.c has /* Standard set of module subdirectories to search, in order of preference */ static const char *stdSubdirs[] = { , input/, drivers/, multimedia/, extensions/, internal/, NULL }; but since a recursive search is done on first, the other directories in the list will be traversed (again and without hope of success) only if the first search was unsuccessful. So that order of preference is no more existant and the above pseudocode could be simplified to modulepath=/usr/lib/xorg/modules search_recursive ${modulepath}/linux/ search_recursive ${modulepath}/ Back to the original problem: replacing libglx.so with a vendor implementation without using diversions. I'd do the following: * leave /usr/lib/xorg/modules/extensions/libglx.so where it is (i.e. no longer divert it) * the proprietary drivers register their libglx.so replacement with the glx alternative which installs it as a slave link to /usr/lib/xorg/modules/linux/libglx.so so that it is picked up first and the original one is being ignored linux is acually OSDIR in the code and may have had a different intention, but it is the only deterministic place that is searched before extensions/libglx.so * no changes in Xorg needed regarding the search paths * works on both unstable and squeeze, so backports don't need special handling If this is OK, I'll implement and test it and also verify that Xorg works in the same way in testing/unstable/experimental. That way, the first of the ugly diversions could be removed. :-) Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4e5373a6.2070...@abeckmann.de
Bug#638965: xserver-xorg-core: add override directory that is prepended to the module search path
Package: xserver-xorg-core Version: 2:1.7.7-13 Severity: wishlist Hi, as this was discussed on the mailing list before, see e.g. [1] http://lists.debian.org/debian-x/2011/08/msg00157.html [2] http://lists.debian.org/debian-x/2011/08/msg00364.html there is need for an override directory that is prepended to the module search path. In [1] I describe my analysis of the current search path system and in [2] Julien says I'd prefer an explicit override directory over abusing 'linux'. An override directory could be used so that vendor implementations of a module (e.g. libglx.so) can be installed installed in the override directory and will be found before the native xorg module without having to use dpkg-divert on the original module. Such a solution would simplify using update-alternatives to enable/disable the vendor implementation without installing/removing packages. For the time being, /usr/lib/xorg/modules/linux/ could be abused for that purpose. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110823113812.26681.77327.reportbug@calzone.localnet
Re: Xorg module searchpath override directory
On 2011-08-23 11:33, Julien Cristau wrote: On Tue, Aug 23, 2011 at 11:32:22 +0200, Andreas Beckmann wrote: Are there any objections to installing the alternative slave link for vendor implemntations of libglx.so as /usr/lib/xorg/modules/linux/libglx.so ... I'd prefer an explicit override directory over abusing 'linux'. OK, just opened wishlist bug #638965 against xserver-xorg-core. http://bugs.debian.org/638965 Until this explicit override directory gets implemented and for squeeze-backports (which needs newer versions of nvidia-graphics-drivers for supporting newer GPUs), would you agree if I use modules/linux for now? That way the first of the diversions can be eliminated and we can go back to the other question How to reorganize the mesa packages to avoid diversions when installing a vendor libGL.so.1. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4e53947c.2080...@abeckmann.de
Xorg module searchpath override directory (was: Re: [RFC] proposal for reorganizing the MESA libraries to simplify replacing them with vendor implementations)
On 2011-07-25 19:03, Julien Cristau wrote: On Mon, Jul 25, 2011 at 18:57:17 +0200, Andreas Beckmann wrote: libGL.so.1 is not the only file to be diverted, libglx.so is being replaced by the vendors drivers, too. Or is there some way of search path magic that would allow to place the higher priority alternative in a different location where it would be picked up before /usr/lib/xorg/modules/extensions/libglx.so ? That would allow to avoid the diversion here. Something similar to /lib/modules/*/updates/ ... That shouldn't be too hard. I tried to look in the code (xorg-server (2:1.7.7-13) hw/xfree86/loader/* ; the experimental version does not differ much in that area) but didn't quite get how the search path is working. Therefore I added a bit debugging code to trace FindModuleInSubdir(). Translating to pseudocode this now looks like modulepath=/usr/lib/xorg/modules for subdir in input drivers multimedia extensions internal do search_recursive ${modulepath}/${subdir}/linux/ search_recursive ${modulepath}/${subdir}/ done So the order the directories are traversed and files are found depends on the raw order in the filesystem (opendir()), except for the linux subdir which is checked first. hw/xfree86/loader/loadmod.c has /* Standard set of module subdirectories to search, in order of preference */ static const char *stdSubdirs[] = { , input/, drivers/, multimedia/, extensions/, internal/, NULL }; but since a recursive search is done on first, the other directories in the list will be traversed (again and without hope of success) only if the first search was unsuccessful. So that order of preference is no more existant and the above pseudocode could be simplified to modulepath=/usr/lib/xorg/modules search_recursive ${modulepath}/linux/ search_recursive ${modulepath}/ Back to the original problem: replacing libglx.so with a vendor implementation without using diversions. I'd do the following: * leave /usr/lib/xorg/modules/extensions/libglx.so where it is (i.e. no longer divert it) * the proprietary drivers register their libglx.so replacement with the glx alternative which installs it as a slave link to /usr/lib/xorg/modules/linux/libglx.so so that it is picked up first and the original one is being ignored linux is acually OSDIR in the code and may have had a different intention, but it is the only deterministic place that is searched before extensions/libglx.so * no changes in Xorg needed regarding the search paths * works on both unstable and squeeze, so backports don't need special handling If this is OK, I'll implement and test it and also verify that Xorg works in the same way in testing/unstable/experimental. That way, the first of the ugly diversions could be removed. :-) Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4e42f368.7030...@abeckmann.de
Re: [RFC] proposal for reorganizing the MESA libraries to simplify replacing them with vendor implementations
On 2011-08-04 11:22, Michel Dänzer wrote: On Mon, 2011-07-25 at 19:03 +0200, Julien Cristau wrote: On Mon, Jul 25, 2011 at 18:57:17 +0200, Andreas Beckmann wrote: libGL.so.1 is not the only file to be diverted, libglx.so is being replaced by the vendors drivers, too. Or is there some way of search path magic that would allow to place the higher priority alternative in a different location where it would be picked up before /usr/lib/xorg/modules/extensions/libglx.so ? That would allow to avoid the diversion here. Something similar to /lib/modules/*/updates/ ... That shouldn't be too hard. But it would result in the non-standard libglx.so being picked up even when the standard GL(X) components are supposed to be used. So I think libglx.so would need to be handled by the alternative as well. The idea is still to handle the non-standard libglx.so by an alternative. This will be installed in the search path before the standard libglx.so. So we can stop diverting the standard libglx.so simplifying the whole process. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4e3a91ff.4080...@abeckmann.de
Re: [RFC] proposal for reorganizing the MESA libraries to simplify replacing them with vendor implementations
[adding pkg-fglrx-devel@ to the discussion] On 2011-07-22 21:12, Cyril Brulebois wrote: Julien Cristau jcris...@debian.org (22/07/2011): So in principle I dislike the idea of making the mesa packages messier to make the closed driver packages' life easier. One thing that's been a source of countless bug in the current system is diversions, because they're evil, and people keep getting them wrong, and users don't understand/expect them, and all kinds of fun ensues. If mesa were to not ship the /usr/lib/$arch/libGL.so.1 (and friends) symlink, but instead ship an alternative itself, would that be enough to put an end to the diversions? Not that I think alternatives are ideal either, but if we're going to have to put up with something I'd rather it wasn't *both* alternatives and diversions. Not sure what other X people think. I'd rather avoid the mess of diversions *AND* alternatives indeed. If adding alternative support to mesa is the price to pay, I guess we could work something out. OK, that is a different solution that should simplify this even more. What is the correct approach to do alternatives of libraries in Multi-Arch: yes packages? * a separate alternative for each architecture (allows weird mixed configurations) * one alternative (from an extra Multi-Arch: foreign package) that covers all arches as slaves links? * the Ubuntu approach of dropping something in ld.so.conf.d/ - would we need one file/one alternative per arch? (I didn't look in detail at the Ubuntu packages, yet.) A gain handling config files with alternatives ... but this time config files that are not meant for user modification (compared to my approach to handle xorg.conf or xorg.conf.d/nvidia.conf with alternatives) libGL.so.1 is not the only file to be diverted, libglx.so is being replaced by the vendors drivers, too. Or is there some way of search path magic that would allow to place the higher priority alternative in a different location where it would be picked up before /usr/lib/xorg/modules/extensions/libglx.so ? That would allow to avoid the diversion here. Something similar to /lib/modules/*/updates/ ... Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4e2da06d.1050...@abeckmann.de
[RFC] proposal for reorganizing the MESA libraries to simplify replacing them with vendor implementations
Hi, currently there are multiple vendor implementations for libGL.so.1/libglx.so and soon there will be vendor implementations for libEGL.so.1/libGLES*.so.*, too. The GL part is currently being handled by diversions and alternatives, and the upcoming EGL part is planning to do this similar. Unfortunately the diversions require quite some knowledge about the files shipped by the MESA packages. Therefore I suggest the following new layout to be used by the MESA packages to simplify diversions and alternatives: /usr/lib[/triplet]/mesa: libFOO.so.1.2.3 libFOO.so.1 - libFOO.so.1.2.3 libFOO.so - libFOO.so.1 /usr/lib[/triplet]: libFOO.so.1 - mesa/libFOO.so.1 libFOO.so - mesa/libFOO.so (for appropriate substitutions of FOO, .1$ , .1.2.3$) Behind this are the following ideas: * move the real file (libFOO.so.1.2.3) out of the scope of ldconfig (which could mess up the SONAME link otherwise) * simplify diverting the libraries: one only needs to know the SONAME (libFOO.so.1) but not the real filename (libFOO.so.1.2.3) which is allowed to change (e.g. libgl1-mesa-swx11 changes the filename with each new upstream release, so it's currently not divertible) * only one diversion per library is needed (/usr/lib[/triplet]/libFOO.so.1) instead of three (libFOO.so.1.2.3, libFOO.so.1, libFOO.so) * the libFOO.so link is not affected by a diversion of libFOO.so.1 - which aides the following behaviour: link with the free mesa library when building the application, but use the accelerated vendor library when running the application * mesa can be provided as a more obvious alternative /usr/lib/mesa I wouldn't limit the new approach to a few architectures (amd64 and i386 need it right now, armel will need it for EGL, soon), as we don't know where vendors will provide custom implementations in the future. I can create a patch for the mesa packages in case someone is interested. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4e285ddc.2000...@abeckmann.de
Re: xorg configuration for the non-free nvidia driver
On 2011-07-11 19:33, Julien Cristau wrote: I very much dislike the idea of using alternatives for configuration files. I understand the concerns regarding alternatives and configuration files, especially if update-alternatives is buggy when it comes to files sitting in the place of alternatives targets (#633627). So does anyone have a suggestion how to simplify the configuration of nvidia-glx using a different approach? This would be interesting mainly for new installations, not updates. * libGL.so.1 and libglx.so need to be replaced by the proprietary versions. This is done automatically by diversions (package glx-diversions) and (slave) alternatives (package glx-alternative- nvidia) upon installation of nvidia-glx. Manual reconfiguration (e.g. switching back to MESA) is easy and does not require package removal: update-alternatives --configure glx * an xorg.conf (or xorg.conf.d/ snippet) has to be created because xorg autoconfiguration does not detect the proprietary drivers, so we need a 'Driver nvidia' line somewhere. Autoconfig of modes etc. usually works fine. The user should be able to add more directives to customize his configuration (e.g. by using nvidia-settings or nvidia-xconfig). An xorg.conf with 'Driver nvidia' breaks Xorg if the proprietary driver is not available, e.g. the glx alternative is set to mesa. So the configuration using the proprietary driver should be moved out of the way in that case. The new apporach should be easily adaptable for other proprietary drivers, e.g. fglrx. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4e1c3a65.5030...@abeckmann.de
Re: xorg configuration for the non-free nvidia driver
On 2011-07-08 20:19, Julien Cristau wrote: On Fri, Jul 8, 2011 at 10:47:50 +0200, Andreas Beckmann wrote: since the proprietary nvidia driver does not work with Xorg autoconfiguration, an xorg.conf is needed to enable it. I'm planning to use debconf for creating a xorg.conf.d snippet on the first installation of xserver-xorg-video-nvidia* (recently split from nvidia-glx*), unless something is already configured manually. I would place this file as /etc/X11/nvidia.conf and use the glx alternatives system to install a slave alternative /etc/X11/xorg.conf.d/nvidia.conf pointing to this file. That way switching to/from nvidia's libGL.so.* and libglx.so would also enable/disable the config file. If this approach works out well for nvidia, I'll propose fglrx to do the Do you have any objections regarding this approach? Yes. I don't think xorg.conf.d should be used for non-InputClass configuration sections at this point. See also https://bugs.freedesktop.org/show_bug.cgi?id=32430 In that case, would it be OK to just install a slave alternative directly as /etc/X11/xorg.conf? The /etc/X11/nvidia.conf would be created only if there was no xorg.conf previously (and no xorg.conf.d/*.conf with a 'Driver .*' line) and the slave link will be created only if nvidia.conf exists. So this should not interfere with systems where an xorg.conf already exists. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4e1b1345.7000...@abeckmann.de
Re: xorg configuration for the non-free nvidia driver
On 2011-07-11 18:26, Russ Allbery wrote: Combining configuration files with slave alternatives makes me nervous. I'm not sure what would happen if the system administrator broke the symlink (via an editor action, for example) while the alternatives system still thinks that it's active; we might end up confusing the alternatives database. Replacing the link with a file is save (at first): update-alternatives: warning: not replacing /etc/X11/xorg.conf.d/nvidia.conf with a link but unfortunately switching to an alternative that does not have this slave causes update-alternatives to delete the file. I'm going to check tonight whether this is a regression from the dpkg in lenny. Replacing the link with a different link would probably be ignored by the alternatives system, e.g. the custom link would be replaced by the alternative link on next update. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4e1b32fd.9010...@abeckmann.de
xorg configuration for the non-free nvidia driver
[resending to debian-x@ instead of just kibi@] Dear X Team, since the proprietary nvidia driver does not work with Xorg autoconfiguration, an xorg.conf is needed to enable it. I'm planning to use debconf for creating a xorg.conf.d snippet on the first installation of xserver-xorg-video-nvidia* (recently split from nvidia-glx*), unless something is already configured manually. I would place this file as /etc/X11/nvidia.conf and use the glx alternatives system to install a slave alternative /etc/X11/xorg.conf.d/nvidia.conf pointing to this file. That way switching to/from nvidia's libGL.so.* and libglx.so would also enable/disable the config file. If this approach works out well for nvidia, I'll propose fglrx to do the same. Do you have any objections regarding this approach? Should I use some ordering prefix for the link installed in /etc/X11/xorg.conf.d, e.g. 42_nvidia.conf or z42_nvidia.conf? Should I use a different filename? Why is /etc/X11/xorg.conf.d/ not shipped by e.g. x11-common? Thanks. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4e16c436.4050...@abeckmann.de
Bug#632886: libgl1-mesa-swx11: drop Conflicts: nvidia-glx
Package: libgl1-mesa-swx11 Version: 7.10.3-3 Severity: wishlist Hi KiBi, libgl1-mesa-swx11 currently has a Conflicts: nvidia-glx entry. While this is correct it's not nearly sufficient (the nvidia-glx package has been split and reorganized, there are legacy packages, too, and there is fglrx as well). With the recent diversions and alternatives changes regarding multi-arch support, all diversion handling has been concentrated in the glx-diversions package. All nvidia and fglrx packages depend on this package (usually indirectly via a longer dependency chain) and glx-diversions now also has proper conflicts with libgl1-mesa-swx11 (which is not divertible since the library file name changes with each new upstream version). Once multiarch aware nvidia-graphics-drivers and glx-alternatives packages enter testing, the users will have less possibilities to break their glx :-) Therefore I suggest you remove this conflict, so you don't have to track current and future possible conflictors. Andreas -- System Information: Debian Release: 6.0.2 APT prefers stable APT policy: (800, 'stable'), (750, 'oldstable'), (700, 'testing'), (600, 'unstable'), (500, 'stable-updates'), (130, 'experimental') Architecture: amd64 (x86_64) Kernel: Linux 2.6.32-5-amd64 (SMP w/2 CPU cores) Locale: LANG=C, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110706190243.32438.55654.reportbug@calzone.localnet
Bug#630710: please add Breaks: libgl1-nvidia-alternatives (= 275.09.07-1)
Package: libgl1-mesa-glx Version: 7.10.2-4 Severity: normal Hi KiBi, could you add Breaks: libgl1-nvidia-alternatives (= 275.09.07-1) to libgl1-mesa-glx? (Best before the multiarch change moves to testing). The multiarch move of MESA breaks current diversion handling (many bugreports). I'm working on updates (but this will have to go through NEW), but for now people could keep working nvidia systems using the Xorg/MESA versions in testing. 275.09.07-1 is a new upstream release to be uploaded soon and will be the last version without a multiarch mesa fix. Thanks Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110616131209.29078.82260.report...@cake.ae.cs.uni-frankfurt.de
Bug#629823: xterm and nvidia
On 2011-06-13 12:04, Stefano Callegari wrote: I have tried more terminal: aterm, eterm and mrxvt. All have the same problem like xterm. So is it a kde4 problem, maybe kwin? Have you tried creating a new user and trying again as that user, so that all KDE user settings get freshly initialized? Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4df8646b.2060...@abeckmann.de
Re: Bug#629823: xterm and nvidia
reassign 629823 xterm 270-1 retitle 629823 can't un-maximize xterm thanks On 2011-06-09 11:47, Stefano Callegari wrote: ... So when konsole is ok and xterm no, I thought that miss something for xterm on new relase of nvidia driver. But today I have tried with more settings (theme, border) and I have the same results: - only with kde4 air-oxygen theme there are some artefacts; This is eventually a separate issue and should be tracked in its own bug report. - with all I can maximize xterm but I can't return to normal dimensions (80x25). I have tested the last above with 2 pc with kde4 and Sid: my with nvidia and one with intel GPU. So I think isn't a nvidia bugs. OK, retitling and reassigning back to xterm. Let's focus on this issue and the non-nvidia setup for now. After maximized xterm, the relate button icon don't change to maximize. It seems that the window is only stretched to max dimensions instead maximized. Before I haven't noted that because I don't use these buttons in my kde4 configuration, but in front the pc with intel GPU I have the surprise ;) Please use reportbug to give followup info to this bug from the intel machine, so that it can recollect system information there. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4df1cf2d.6030...@abeckmann.de
Re: Diversions
On 2011-02-16 18:53, Cyril Brulebois wrote: Hi guys, could you please tell us which files you divert in (all of) your nvidia packages? I'd like to get those reported in the bug script, so either file lists or patterns would be appreciated. We divert /usr/lib/libGL.so* /usr/lib32/libGL.so* /usr/lib/xorg/modules/extensions/libglx.so In the nvidia-glx bug-script (in SVN) I use the following to list the interesting stuff about the links, diversions and alternatives: cat EOF 3 OpenGL and NVIDIA library files installed: `ls -la /etc/alternatives/*libGL* /usr/lib/libGL.* /usr/lib/libGLcore* /usr/lib/libnvidia* /usr/lib32/libGL.* /usr/lib32/libGLcore* /usr/lib32/libnvidia* 2/dev/null` `ls -la /usr/lib/nvidia /usr/lib/nvidia/diversions /usr/lib32/nvidia /usr/lib32/nvidia/diversions 2/dev/null` Files from nvidia-installer: `ls -la /usr/bin/nvidia-installer /usr/bin/nvidia-uninstall /var/lib/nvidia 2/dev/null` EOF Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4d5c11d5.7090...@abeckmann.de
Re: Bug#610021/#610022: fglrx-glx fails to install in parallel with nvidia-glx
On Sunday, 16. January 2011 20:24:17 Patrick Matthäi wrote: Am 16.01.2011 14:49, schrieb Ronny Standtke: reopen 610022 Sorry, not possible This is not true, it is definitely possible. Several parties have already developed solutions for this problem, see the following both messages of the discussion on the Debian Live mailinglist mentioned in the initial bug report: http://lists.debian.org/debian-live/2011/01/msg00065.html http://lists.debian.org/debian-live/2011/01/msg00066.html Let me describe how the new nvidia non-free packages work for squeeze: (I implemented and tested these changes.) There are three new packages libgl1-nvidia-alternatives, libgl1-nvidia-alternatives-ia32, libgl1-nvidia-alternatives-ia32 that do the diversions of OpenGL libraries (from libgl1-mesa-glx, ia32-libs, *-dev) and create alternatives of the diverted files. The diversions are created/removed by dpkg triggers whenever the diverted packages are (un-)installed. The nvidia packages Depends: on the lib*-nvidia-alternatives package for the library they provide, install the new library in a private library directory (/usr/lib/nvidia) and add an alternative (with higher priority than the diverted library from mesa). There are currently a lot of Conflicts defined to make sure only one of nvidia-graphics-drivers, nvidia-graphics-drivers-legacy-173xx or nvidia-graphics-drivers-legacy-96xx can be installed at a time. (legacy-71xx is no longer supported with current Xorg.) Most of these Conflicts should not be necessary since there are no more file name conflicts for the libraries. There are file name conflicts for nvidia_drv.so (the Xorg driver module) libglx.so (the replaced Xorg module) and nvidia.ko (the kernel module which may exist for any number of distribution and custom kernels at the same time) and I currently have no solution how to solve these to allow installing both nvidia-graphics-drivers as well as one (or two) legacy versions at the same time. Could we rename them eventually? Alternatively (at least for the xorg modules) we could set up one more set of alternatives. ... About configuring xorg for the nonfree drivers: There is a compile time option for Xorg that makes it read PCI ID lists from /usr/share/xserver-xorg/pci/ and auto-select drivers modules of corresponding file names if a matching device was found in the system. This flag was active for some tome in the Debian Xorg packages, but is no longer enabled. IIRC it was related to defining PCI_TXT_IDS_DIR. The nvidia-glx* packages ship a nvidia.ids file in that directory (but it's not used by current xorg). So by installing fglrx it is also not possible to use another 3d accelerated gpu driver anymore. It is, you just need to use update-alternative. And it does work for nvidia-glx :-) The xorg/mesa team has to play with us then. I also don't think, that using the alternative system is the right way, because: 1) building against the fglrx libgl implementation is a bad idea Agreed. Therefore nvidia-graphics-drivers* does not provide any libGL.so symlink. Instead the libgl1-nvidia-alternatives package diverts the libGL.so link provided by libgl1-mesa-dev and creates an alternative (via triggers) pointing to the diverted libGL.so - so at link time always the mesa opengl library will be used. For picking up the correct library at package build time, a shlibs file redirecting to libgl1-mesa-glx is being used. (I.e. in short - no building/packaging of opengl programs without libgl1-mesa-dev installed). 2) users may get strange problems/crashes, if they install different drivers (fglrx+free-drivers / nvidia+fglrx / nvidia+free drivers etc). Packages and alternatives should cooperate and have safe defaults - and bug-scripts that list all possible breakers :-) 3) It is no alternative, it is a diversion, because it is needed! It is alternatives. The diversions are only needed until we can convince the mesa maintainers to ship their library in /usr/lib/mesa and install a primary alternative. (I haven't tried and I'm not sure if we really want to propose this). 4) for fglrx it would just be a little step for live booting, fglrx still needs a xorg.conf with some options (like defaultdepth) nvidia-glx currently needs the following minimal xorg.conf to be used (since there is no autodetection) Section Device Identifier GPU Driver nvidia EndSection I understand your position and it is a good idea to add fglrx to live cds, but I don't see any good solution (without ugly workarounds) to implement it, yet. Something I can offer to the fglrx maintainers is to generalize the diversion/alternatives packages so that it supports fglrx, too. This should be not too difficult. But eventually we should rename the packages (something that contains xorg and non-free perhaps?) and move the location of the diversions to remove the references to nvidia. I can also help getting the
Autoconfiguring third-party Xorg drivers
tags 586502 - patch clone 586502 -1 retitle -1 Please reenable PCI_TXT_IDS_DIR for third-party drivers submitter -1 ! severity -1 wishlist reassign -1 xserver-xorg-core 2:1.7.7-4 thanks Dear Debian X Strike Force, please reenable the PCI_TXT_IDS_DIR feature of Xorg. Even if it is not used by Xorg itself, it would simplify automatic configuration of third party drivers (the non-free nvidia driver in this case). Xorg does not (and needs not to) know details about the existence of these drivers as long as finding a matching PCI ID redirects it to the correct driver, Currently an xorg.conf has to be created by the user because autoconfiguration does not use the proprietary driver. Or is there another solution to plug some driver into Xorg without creating an xorg.conf? Thanks. Andreas -- To UNSUBSCRIBE, email to debian-x-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4c747d11.5090...@abeckmann.de
Bug#466526: xorg starts with an oversized xrandr screen
Hi Brice, Brice Goglin wrote: Andreas Beckmann wrote: Package: xorg Version: 1:7.3+10 Severity: normal Hi, if xorg detects a second monitor connected, it starts up with a virtual screen larger than what I set in xorg.conf. (X700, laptop 1680x1050 right-of flat panel @ vga port 1600x1200) Output of xrandr after startup, note that current is larger than maximum: Is it the same if you try to put the LVDS on the left on the VGA monitor? No, in this case it seems to work fine. But the original bug is still reproducible in the VGA-leftof-LVDS setup with the latest xserver-xorg-video-ati from experimental. What if you remove the xorg.conf Right-Of stuff and just call xrandr? You never get the oversized screen right? (I just want to be sure that only the xorg.conf stuff is broken, and that runtime xrandr changes are ok). It starts with an 1680x1200 screen (which does not fit on any of the connected screens) and then I have no idea how I could trick xrandr to invalid-resize the screen. Running xrandr gives a correct dual head setup (provided there is a sufficient Screen/Display/Virtual setting, e.g. '3280 1200', otherwise resizing fails with an explanatory error message). But there is a new bug: If there is no LeftOf/RightOf setting in xorg.conf and I use xrandr to setup one (or I use xrandr to switch the setting from xorg.conf to the opposite layout) the following happens: maximizing a window does not maximize it to the size of the current subscreen (either LVDS or VGA-0) but to either total-width-of-virtual-screen x height-of-current-subscreen i.e. spanning both subscreens (no LeftOf/RightOf in xorg.conf) or initial-dimensions-of-the-subscreen i.e. does not look maximized (RightOf in xorg.conf switched to LeftOf with xrandr) Should I report this as a separate bug? I can make screenshots etc. Make sure you try the latest xserver-xorg-video-ati from experimental and xserver-xorg-core from unstable. If they don't help, I will forward this bug upstream. Above tests were done with these versions: ii xserver-xorg-video-ati 1:6.8.1~git20080320.5e3b2128-1 ii xserver-xorg-core 2:1.4.1~git20080131-2 Andreas PS: something I didn't test: The LeftOf/RightOf setting in xorg.conf can be done for both LVDS and VGA-0, I only tested it with VGA-0. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#466526: more on oversized xrandr screen
When I increase the size of the virtual screen to 3360x1200, it uses this size at startup (as before, but it's no longer 'oversized') and there is no distorted image at the left and right border. There were probably 80 pixel overlapping at both sides before. Andreas -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#460269: xkb-data: layout=us + variant=nodeadkeys breaks Ctrl+F1 ...
Package: xkb-data Version: 1.0~cvs.20070916-1 Severity: normal Hi, when I recently switched the x keyboard config from layout=de to layout=us, this broke Ctrl+F1, ... i.e. switching to the ttys which didn't react at all. I checked a bit further and found this to be caused by the Option XkbVariantnodeadkeys which was left from the previous config. After removing this, switching to the ttys worked again. Andreas -- System Information: Debian Release: lenny/sid APT prefers testing APT policy: (500, 'testing'), (500, 'stable'), (300, 'unstable'), (130, 'experimental') Architecture: i386 (i686) Kernel: Linux 2.6.23-1-686 (SMP w/1 CPU core) Locale: LANG=C, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/bash -- no debconf information -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]