Created attachment 99820 [details] GDB session log The assert hit is very occasional/random. Sometimes it is reproducible several times in row, sometimes several days go on without problems. Attaching a GDB log with (some of) the information requested in IRC by Chris Wilson. Unfortunately not all variables requested by Chris could be printed out, due to having been optimized out by compiler. -- Window manager: WindowMaker 0.95.5 -- chipset: 00:02.0 VGA compatible controller: Intel Corporation 82G33/G31 Express Integrated Graphics Controller (rev 10) -- system architecture: i686 / 32bit -- xf86-video-intel: GIT 605fcd9050efc816ac8163e8d626f466d98261c2 -- xserver: X.Org X Server 1.15.1-1 from latest Debian testing -- mesa: 10.1.2-1 from Debian testing -- libpixman: 0.32.4-1 -- libdrm version: 2.4.54-1 -- kernel version: 3.14.4 (vanilla+grsec) -- Linux distribution: current Debian Testing -- Machine or mobo model: Asus P5KPL-CM -- Display connector: VGA
An io bo, at least that reduces the amount of code to think about first. (gdb) p *src_bo $1 = {rq = 0xb7cfc358, exec = 0x0, proxy = 0xb81315e8, list = {next = 0xb7e150e4, prev = 0xb7e150e4}, request = {next = 0xb7e150ec, prev = 0xb7e150ec}, vma = {next = 0xb7e150f4, prev = 0xb7e150f4}, map__cpu = 0xb3a2d000, map__gtt = 0x0, binding = {next = 0x0, format = 0, offset = 0}, presumed_offset = 0, unique_id = 3765653, refcnt = 0, handle = 176, target_handle = 4294967295, delta = 4096, size = {pages = {count = 224, bucket = 0}, bytes = 224}, pitch = 56, tiling = 0, reusable = 0, gpu_dirty = 0, gtt_dirty = 0, domain = 0, needs_flush = 0, snoop = 0, io = 1, flush = 0, scanout = 0, purged = 0}
Created attachment 99826 [details] [review] Assert we do not replace live IO buffers
Created attachment 99827 [details] [review] Implicit release of upload buffers considered bad A couple of patches to test. Both address 2 little bugs, the second is more likely to explain your assertion failures.
Created attachment 99835 [details] [review] Implicit release of upload buffers considered bad
Created attachment 99836 [details] [review] Implicit release of upload buffers considered bad
Went ahead and pushed: commit 9f4f855ba37966fb91d31e9081d03cf72affb154 Author: Chris Wilson <chris@chris-wilson.co.uk> Date: Mon May 26 07:06:18 2014 +0100 sna: Implicit release of upload buffers considered bad Currently upload buffers are automatically decoupled when the buffer is retired. As retiring can happen during command setup after we have selected which bo to render with, this can free the bo we plan to use. Which is bad. Instead of making the release of upload buffers automatic, we manually check whether the buffer is idle before use as a source to consider scrapping it and replacing it with a real GPU bo. This is likely to keep upload buffers alive for longer (limiting reuse between Pixmaps but making reuse of the buffer within a Pixmap more likely) which is both good and bad. (Good - may improve the content cache, bad - may increase the amount of memory used by upload buffers for arbitrary long periods.) Reported-by: Matti Hämäläinen <ccr@tnsp.org> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=79238 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> commit b508d8f3318c42a2a87b7731789b1d03610e9b46 Author: Chris Wilson <chris@chris-wilson.co.uk> Date: Mon May 26 06:54:03 2014 +0100 sna: Assert that we do not replace active IO buffers References: https://bugs.freedesktop.org/show_bug.cgi?id=79238 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Thanks, will be testing this.
The good news is that after ~two weeks of testing, also with the later changes mentioned on IRC, I've encountered this issue no more. I'd say with rather good confidence that the issue can be closed. Thanks again for your hard work!
Thanks for the testing and feedback.
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.