<html> <head> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> </head> <body text="#000000" bgcolor="#FFFFFF"> Thanks for fix. Regards, David Zhou <div class="moz-cite-prefix">On 2018年02月19日 15:50, Christian König wrote: </div> <blockquote type="cite" cite="mid:032f5418-4b01-f269-7e0a-5e1dcad65e10@gmail.com"> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> <div class="moz-cite-prefix">Patch is Acked-by: Christian König <a class="moz-txt-link-rfc2396E" href="mailto:christian.koenig@amd.com" moz-do-not-send="true"><christian.koenig@amd.com></a> Regards, Christian. Am 19.02.2018 um 08:25 schrieb Andrey Grodzovsky: </div> <blockquote type="cite" cite="mid:dcb5a4b1-f959-107b-d4af-18758fbf74cd@amd.com">Found time to do some debugging and found error in assigning of initial VA hole size, attached patch makes this issue go away, otherwise amdgpu_vamgr_find_va would always fail. Thanks, Andrey On 02/18/2018 03:41 AM, Andrey Grodzovsky wrote: <blockquote type="cite">Hi David, I have an issue on Vega 10 where sudo xinit fails with [ 5801.227] Failed to allocate front buffer memory [ 5801.227] (EE) AMDGPU(0): amdgpu_setup_kernel_mem failed [ 5801.227] (EE) Fatal server error: [ 5801.227] (EE) AddScreen/ScreenInit failed for driver 0 [ 5801.227] (EE) [ 5801.227] (EE) Also in MESA code si_create_context->si_common_context_init->pipe_buffer_create returns NULL. Ellsmire works fine, on Vega 10 reverting this patch makes it work. I am during my test system upgrade right now with some more issues after this one, so I can spend time on debugging this once my system is in good shape, unless you have an answer to this problem right away. Thanks, Andrey On 02/09/2018 03:58 AM, Michel Dänzer wrote: <blockquote type="cite">On 2018-02-09 07:01 AM, Chunming Zhou wrote: <blockquote type="cite">On 2018年02月08日 17:13, Chunming Zhou wrote: <blockquote type="cite">On 2018年02月08日 17:01, Michel Dänzer wrote: <blockquote type="cite">this change completely broke radeonsi due to memory management errors (see valgrind output for glxgears below), so I had to revert it. </blockquote> ok, I will look into what happens. Sorry for broken. </blockquote> The patch lost a "goto out" in free_va when applying patch from hybrid to opensource. The attached fixes that, I tested successfully in my local side. please review or test, if no problem, I will submit it again. </blockquote> Please add: v2: Add missing "goto out" Tested-by: Michel Dänzer <a class="moz-txt-link-rfc2396E" href="mailto:michel.daenzer@amd.com" moz-do-not-send="true"><michel.daenzer@amd.com></a> to the commit log before pushing. </blockquote> _______________________________________________ amd-gfx mailing list <a class="moz-txt-link-abbreviated" href="mailto:amd-gfx@lists.freedesktop.org" moz-do-not-send="true">amd-gfx@lists.freedesktop.org</a> <a class="moz-txt-link-freetext" href="https://lists.freedesktop.org/mailman/listinfo/amd-gfx" moz-do-not-send="true">https://lists.freedesktop.org/mailman/listinfo/amd-gfx</a> </blockquote> <fieldset class="mimeAttachmentHeader"></fieldset> <pre wrap="">_______________________________________________ amd-gfx mailing list <a class="moz-txt-link-abbreviated" href="mailto:amd-gfx@lists.freedesktop.org" moz-do-not-send="true">amd-gfx@lists.freedesktop.org</a> <a class="moz-txt-link-freetext" href="https://lists.freedesktop.org/mailman/listinfo/amd-gfx" moz-do-not-send="true">https://lists.freedesktop.org/mailman/listinfo/amd-gfx</a> </pre> </blockquote> </blockquote> </body> </html>