Summary: | no effective checking pciaccess version when compiling with xserver-1.3 | ||
---|---|---|---|
Product: | xorg | Reporter: | Sérgio M. Basto <sergio> |
Component: | Driver/intel | Assignee: | Eric Anholt <eric> |
Status: | RESOLVED INVALID | QA Contact: | Xorg Project Team <xorg-team> |
Severity: | normal | ||
Priority: | medium | CC: | dri-devel, jcnengel, nian.wu, shuang.he |
Version: | git | ||
Hardware: | Other | ||
OS: | All | ||
Whiteboard: | |||
i915 platform: | i915 features: |
Hi, after many work , I put my system with xorg-server-1.4 and all Xorg 7.4, and finally I been able to compile latest xor-x11-drv.intel . I think the drive should warning us to this fact. with xorg-server 1.3: checking whether XSERVER_LIBPCIACCESS is declared... no checking for PCIACCESS... yes and badly pass with xor-server 1.4: checking whether XSERVER_LIBPCIACCESS is declared... yes checking for PCIACCESS... configure: error: Package requirements (pciaccess >= 0.10.0) were not met: Requested 'pciaccess >= 0.10.0' but version of pciaccess is 0.9.1 (In reply to comment #2) > Is there any relation to bug #5714? > I don't know, may be in this bug , this merge require xorg-server >= 1.4 comment #1 - yes, you do need updated libpciaccess. would you please upgrade pciaccess lib and re-test? (In reply to comment #5) > would you please upgrade pciaccess lib and re-test? > I think, I don't explain my self correctly. The problem is with xorg-server 1.3, no matter the version of LIBPCIACCESS the test pass and perhaps shouldn't pass ( checking whether XSERVER_LIBPCIACCESS is declared... no checking for PCIACCESS... yes ) I'm a bit of confused , too... so let's get back to the beginning of the question. Is the bug about "glxinfo failure" or "libpciaccess dependency issue during compiling"?
>Hi, after many work , I put my system with xorg-server-1.4 and all Xorg 7.4,
>and finally I been able to compile latest xor-x11-drv.intel .
And, you still can't successfully build the intel driver in this environment, right?
Did you update kernel module "drm" and libdrm? (In reply to comment #7) > >Hi, after many work , I put my system with xorg-server-1.4 and all Xorg 7.4, > >and finally I been able to compile latest xor-x11-drv.intel . > > And, you still can't successfully build the intel driver in this environment, > right? No , Finally I build xorg-x11-drv.intel , with xorg-server-1.4 and git of libpciaccess > I'm a bit of confused , too... so let's get back to the beginning of the > question. Is the bug about "glxinfo failure" or "libpciaccess dependency issue > during compiling"? both, may be we can close and forget this bug, because this was when I git pull drm, libdrm, mesa and drv-video-intel with xorg-x11-server 1.3 . The only thing that remains, in this bug, for me, is the fact that drv-video-intel requires a pciaccess >= 0.10.0 with xorg-x11-server >= 1.4 and don't require with xorg-x11-server <= 1.3 , which could be a little bug. What I want is the autotools also require pciaccess >= 0.10.0 for older xorg-x11-server, if it require ok ? else forget it please , and sorry for the noise . if we know that new intel drive needs a pciaccess >= 0.10.0 and xorg-x11-server <=1.3 use a version of pciaccess < 0.10.0 , configuration of xf86-drv-intel should stop and require update or xorg-x11-server or pciaccess. that the point on this bug ok ? I'm changing the title to reflect the pciaccess version issue. Eric, xserver-1.3 is supposed to work with the intel driver git, right? If so, it makes sense to improve the checking mechanism. It is not a bug that new libpciaccess is only required for new server. We can get along fine with old libpciaccess for tools if we don't need libpciaccess for the driver. |
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.
Hi, since commit 126031655c3c1db9881ad17d712461533eb59632 , my i915GM don't work and recent git stop working at all. when i915tex has been moved to i915 . I got on glxinfo: libGL error: drmMap of framebuffer failed (Invalid argument) libGL error: reverting to (slow) indirect rendering display: :0 screen: 0 direct rendering: No (If you want to find out why, try setting LIBGL_DEBUG=verbose) And I don't have DRI with Xorg.0.log: > (WW) intel(0): Removed DRI frontbuffer mapping in compatibility mode. > (WW) intel(0): DRIGetDeviceInfo will report incorrect frontbuffer > handle. Now with latest intel drv git X don't startup : once I got this errors: Oct 11 22:39:35 localhost console-kit-daemon[1979]: GLib-CRITICAL: g_async_queue_push: assertion `g_atomic_int_get (&queue->ref_count) > 0' failed Oct 11 22:39:37 localhost console-kit-daemon[1979]: GLib-CRITICAL: g_async_queue_lock: assertion `g_atomic_int_get (&queue->ref_count) > 0' failed Oct 11 22:39:37 localhost console-kit-daemon[1979]: GLib-CRITICAL: g_async_queue_length_unlocked: assertion `g_atomic_int_get (&queue->ref_count) > 0' failed Oct 11 22:39:37 localhost kernel: console-kit-dae[1987]: segfault at 39312d34 eip 002241d7 esp b7ed3340 error 4 but now I don't have errors neither X startup For me the most strange is nobody complains but me . Have we something new at building drivers that I am missing ?