<!DOCTYPE html> <html> <head> <meta http-equiv="Content-Type" content="text/html; charset=UTF-8"> </head> <body> <div class="moz-cite-prefix">On 12/11/2024 4:24 PM, Cavitt, Jonathan wrote: </div> <blockquote type="cite" cite="mid:CH0PR11MB5444CF26B806C6784B54AEFBE53E2@CH0PR11MB5444.namprd11.prod.outlook.com"> <meta http-equiv="Content-Type" content="text/html; charset=UTF-8"> <meta name="Generator" content="Microsoft Word 15 (filtered medium)"> <style>@font-face {font-family:"Cambria Math"; panose-1:2 4 5 3 5 4 6 3 2 4;}@font-face {font-family:Calibri; panose-1:2 15 5 2 2 2 4 3 2 4;}@font-face {font-family:Aptos;}@font-face {font-family:Consolas; panose-1:2 11 6 9 2 2 4 3 2 4;}p.MsoNormal, li.MsoNormal, div.MsoNormal {margin:0in; font-size:12.0pt; font-family:"Aptos",sans-serif;}a:link, span.MsoHyperlink {mso-style-priority:99; color:blue; text-decoration:underline;}pre {mso-style-priority:99; mso-style-link:"HTML Preformatted Char"; margin:0in; font-size:10.0pt; font-family:"Courier New";}span.HTMLPreformattedChar {mso-style-name:"HTML Preformatted Char"; mso-style-priority:99; mso-style-link:"HTML Preformatted"; font-family:Consolas;}span.hljs {mso-style-name:hljs;}span.EmailStyle22 {mso-style-type:personal-reply; font-family:"Aptos",sans-serif; color:windowtext;}.MsoChpDefault {mso-style-type:export-only; font-size:10.0pt; mso-ligatures:none;}div.WordSection1 {page:WordSection1;}</style> <div class="WordSection1"> <o:p> </o:p> <o:p> </o:p> <div> <div style="border:none;border-top:solid #E1E1E1 1.0pt;padding:3.0pt 0in 0in 0in"> From: Nirmoy Das <a class="moz-txt-link-rfc2396E" href="mailto:nirmoy.das@linux.intel.com"><nirmoy.das@linux.intel.com></a> Sent: Friday, December 6, 2024 11:28 PM To: Cavitt, Jonathan <a class="moz-txt-link-rfc2396E" href="mailto:jonathan.cavitt@intel.com"><jonathan.cavitt@intel.com></a>; <a class="moz-txt-link-abbreviated" href="mailto:intel-xe@lists.freedesktop.org">intel-xe@lists.freedesktop.org</a> Cc: Gupta, saurabhg <a class="moz-txt-link-rfc2396E" href="mailto:saurabhg.gupta@intel.com"><saurabhg.gupta@intel.com></a>; Zuo, Alex <a class="moz-txt-link-rfc2396E" href="mailto:alex.zuo@intel.com"><alex.zuo@intel.com></a>; Vivi, Rodrigo <a class="moz-txt-link-rfc2396E" href="mailto:rodrigo.vivi@intel.com"><rodrigo.vivi@intel.com></a> Subject: Re: [PATCH v2] drm/xe/xe_sync: Add debug printing to check_ufence<o:p></o:p> </div> </div> <o:p> </o:p> <o:p> </o:p> <div> On 12/6/2024 7:11 PM, Jonathan Cavitt wrote:<o:p></o:p> </div> <blockquote style="margin-top:5.0pt;margin-bottom:5.0pt"> <pre style="margin-left:.5in">The xe_sync helper function check_ufence can occasionally report EBUSY<o:p></o:p></pre> <pre style="margin-left:.5in">if the ufence has not been signalled yet. EBUSY is a non-fatal error<o:p></o:p></pre> <pre style="margin-left:.5in">value for the function, so it is not desireable to warn in cases where<o:p></o:p></pre> <pre style="margin-left:.5in">EBUSY is reported because it is up to the user to decide if EBUSY is a<o:p></o:p></pre> <pre style="margin-left:.5in">fatal error in their use cases. However, we can and should report EBUSY<o:p></o:p></pre> <pre style="margin-left:.5in">to the debug logs for diagnostic purposes.<o:p></o:p></pre> <pre style="margin-left:.5in"><o:p> </o:p></pre> <pre style="margin-left:.5in">v2: Use vm_dbg instead of XE_IOCTL_DBG (Rodrigo)<o:p></o:p></pre> <pre style="margin-left:.5in"><o:p> </o:p></pre> <pre style="margin-left:.5in">Signed-off-by: Jonathan Cavitt <a href="mailto:jonathan.cavitt@intel.com" moz-do-not-send="true"><jonathan.cavitt@intel.com></a><o:p></o:p></pre> <pre style="margin-left:.5in">CC: Rodrigo Vivi <a href="mailto:rodrigo.vivi@intel.com" moz-do-not-send="true"><rodrigo.vivi@intel.com></a><o:p></o:p></pre> <pre style="margin-left:.5in">---<o:p></o:p></pre> <pre style="margin-left:.5in"> drivers/gpu/drm/xe/xe_vm.c | 10 ++++++++--<o:p></o:p></pre> <pre style="margin-left:.5in"> 1 file changed, 8 insertions(+), 2 deletions(-)<o:p></o:p></pre> <pre style="margin-left:.5in"><o:p> </o:p></pre> <pre style="margin-left:.5in">diff --git a/drivers/gpu/drm/xe/xe_vm.c b/drivers/gpu/drm/xe/xe_vm.c<o:p></o:p></pre> <pre style="margin-left:.5in">index 74d684708b00..8c770d1b916c 100644<o:p></o:p></pre> <pre style="margin-left:.5in">--- a/drivers/gpu/drm/xe/xe_vm.c<o:p></o:p></pre> <pre style="margin-left:.5in">+++ b/drivers/gpu/drm/xe/xe_vm.c<o:p></o:p></pre> <pre style="margin-left:.5in">@@ -2402,8 +2402,11 @@ static int op_lock_and_prep(struct drm_exec *exec, struct xe_vm *vm,<o:p></o:p></pre> <pre style="margin-left:.5in"> break;<o:p></o:p></pre> <pre style="margin-left:.5in"> case DRM_GPUVA_OP_REMAP:<o:p></o:p></pre> <pre style="margin-left:.5in"> err = check_ufence(gpuva_to_vma(op->base.remap.unmap->va));<o:p></o:p></pre> <pre style="margin-left:.5in">- if (err)<o:p></o:p></pre> <pre style="margin-left:.5in">+ if (err) {<o:p></o:p></pre> <pre style="margin-left:.5in">+ vm_dbg(&vm->xe->drm,<o:p></o:p></pre> <pre style="margin-left:.5in">+ "REMAP: vma check ufence status = %i\n", err);<o:p></o:p></pre> </blockquote> Move that to check_ufence() instead so there there is only one copy of logging ? <o:p></o:p> IMO I think there’s value in knowing which operation is failing, and that information would<o:p></o:p> be lost if we moved the logging into check_ufence. </div> </blockquote> igt stack strace[1] dumps the operation info and any UMD should be able know the failed operation but not against as it is now: Reviewed-by: Nirmoy Das <a class="moz-txt-link-rfc2396E" href="mailto:nirmoy.das@intel.com"><nirmoy.das@intel.com></a> [1] <a class="moz-txt-link-freetext" href="https://intel-gfx-ci.01.org/tree/intel-xe/IGTPW_12052/shard-bmg-8/igt@xe_exec_compute_mode@many-rebind.html">https://intel-gfx-ci.01.org/tree/intel-xe/IGTPW_12052/shard-bmg-8/igt@xe_exec_compute_mode@many-rebind.html</a> <blockquote type="cite" cite="mid:CH0PR11MB5444CF26B806C6784B54AEFBE53E2@CH0PR11MB5444.namprd11.prod.outlook.com"> <div class="WordSection1"> <o:p></o:p> -Jonathan Cavitt<o:p></o:p> Regards,<o:p></o:p> Nirmoy <o:p></o:p> <o:p></o:p> <blockquote style="margin-top:5.0pt;margin-bottom:5.0pt"> <pre style="margin-left:.5in"><o:p> </o:p></pre> <pre style="margin-left:.5in"> break;<o:p></o:p></pre> <pre style="margin-left:.5in">+ }<o:p></o:p></pre> <pre style="margin-left:.5in"> <o:p></o:p></pre> <pre style="margin-left:.5in"> err = vma_lock_and_validate(exec,<o:p></o:p></pre> <pre style="margin-left:.5in"> gpuva_to_vma(op->base.remap.unmap->va),<o:p></o:p></pre> <pre style="margin-left:.5in">@@ -2415,8 +2418,11 @@ static int op_lock_and_prep(struct drm_exec *exec, struct xe_vm *vm,<o:p></o:p></pre> <pre style="margin-left:.5in"> break;<o:p></o:p></pre> <pre style="margin-left:.5in"> case DRM_GPUVA_OP_UNMAP:<o:p></o:p></pre> <pre style="margin-left:.5in"> err = check_ufence(gpuva_to_vma(op->base.unmap.va));<o:p></o:p></pre> <pre style="margin-left:.5in">- if (err)<o:p></o:p></pre> <pre style="margin-left:.5in">+ if (err) {<o:p></o:p></pre> <pre style="margin-left:.5in">+ vm_dbg(&vm->xe->drm,<o:p></o:p></pre> <pre style="margin-left:.5in">+ "UNMAP: vma check ufence status = %i\n", err);<o:p></o:p></pre> <pre style="margin-left:.5in"> break;<o:p></o:p></pre> <pre style="margin-left:.5in">+ }<o:p></o:p></pre> <pre style="margin-left:.5in"> <o:p></o:p></pre> <pre style="margin-left:.5in"> err = vma_lock_and_validate(exec,<o:p></o:p></pre> <pre style="margin-left:.5in"> gpuva_to_vma(op->base.unmap.va),<o:p></o:p></pre> </blockquote> </div> </blockquote> </body> </html>