<html> <head> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> </head> <body text="#000000" bgcolor="#FFFFFF"> int amdgpu_vce_ring_test_ring(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; uint32_t rptr = amdgpu_ring_get_rptr(ring); unsigned i; int r, timeout = adev->usec_timeout; /* skip ring test for sriov*/ if (amdgpu_sriov_vf(adev)) return 0; r = amdgpu_ring_alloc(ring, 16); if (r) return r; amdgpu_ring_write(ring, VCE_CMD_END); amdgpu_ring_commit(ring); Above is original code, rptr is updated when called, and below is your patch, my question is why do you need to get rptr twice? <pre class="moz-quote-pre" wrap="">@@ -1084,6 +1084,8 @@ int amdgpu_vce_ring_test_ring(struct amdgpu_ring *ring) if (r) return r; + rptr = amdgpu_ring_get_rptr(ring); + amdgpu_ring_write(ring, VCE_CMD_END); amdgpu_ring_commit(ring); </pre> <div class="moz-cite-prefix">On 5/27/19 12:22 PM, Li, Ching-shih (Louis) wrote: </div> <blockquote type="cite" cite="mid:MN2PR12MB34398CEE701707E25781E9A2AA1D0@MN2PR12MB3439.namprd12.prod.outlook.com"> <meta name="Generator" content="Microsoft Word 15 (filtered medium)"> <style></style> <div class="WordSection1"> Hi Leo,<o:p></o:p> <o:p> </o:p> Yes, I confirm it is the root cause for the Chrome S3 issue. Whenever system is resumed, the original instruction always gets zero. However, I have no idea why it fails, and didn’t verify this problem on CRB or any other Linux platform yet.<o:p></o:p> Although I think the ideal solution is an indicator, e.g. a register, for driver to check if related firmware and hardware are ready to work. So driver can make sure it is ok to read rptr. Without any reference document, I can only try to solve the problem by modifying driver. Debug traces reveal that only first rptr read fails, but the read in check loop is ok. Therefore, a solution comes to mind: to update rptr later for initial rptr value. Tests prove it working in Chrome platforms. Fyi~<o:p></o:p> <o:p> </o:p> BR,<o:p></o:p> Louis<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: Liu, Leo <a class="moz-txt-link-rfc2396E" href="mailto:Leo.Liu@amd.com"><Leo.Liu@amd.com></a> Sent: Monday, May 27, 2019 9:01 PM To: S, Shirish <a class="moz-txt-link-rfc2396E" href="mailto:Shirish.S@amd.com"> <Shirish.S@amd.com></a>; Grodzovsky, Andrey <a class="moz-txt-link-rfc2396E" href="mailto:Andrey.Grodzovsky@amd.com"> <Andrey.Grodzovsky@amd.com></a>; Zhang, Jerry <a class="moz-txt-link-rfc2396E" href="mailto:Jerry.Zhang@amd.com"> <Jerry.Zhang@amd.com></a>; Deng, Emily <a class="moz-txt-link-rfc2396E" href="mailto:Emily.Deng@amd.com"> <Emily.Deng@amd.com></a>; Deucher, Alexander <a class="moz-txt-link-rfc2396E" href="mailto:Alexander.Deucher@amd.com"> <Alexander.Deucher@amd.com></a> Cc: <a class="moz-txt-link-abbreviated" href="mailto:amd-gfx@lists.freedesktop.org"> amd-gfx@lists.freedesktop.org</a>; Li, Ching-shih (Louis) <a class="moz-txt-link-rfc2396E" href="mailto:Ching-shih.Li@amd.com"> <Ching-shih.Li@amd.com></a> Subject: Re: [PATCH] drm/amdgpu: fix ring test failure issue during s3 in vce 3.0<o:p></o:p> </div> </div> <o:p> </o:p> <o:p> </o:p> <div> On 5/27/19 3:42 AM, S, Shirish wrote:<o:p></o:p> </div> <blockquote style="margin-top:5.0pt;margin-bottom:5.0pt"> <pre>From: Louis Li <a href="mailto:Ching-shih.Li@amd.com" moz-do-not-send="true"><Ching-shih.Li@amd.com></a><o:p></o:p></pre> <pre><o:p> </o:p></pre> <pre>[What]<o:p></o:p></pre> <pre>vce ring test fails consistently during resume in s3 cycle, due to<o:p></o:p></pre> <pre>mismatch read & write pointers.<o:p></o:p></pre> <pre>On debug/analysis its found that rptr to be compared is not being<o:p></o:p></pre> <pre>correctly updated/read, which leads to this failure.<o:p></o:p></pre> <pre>Below is the failure signature:<o:p></o:p></pre> <pre> [drm:amdgpu_vce_ring_test_ring] *ERROR* amdgpu: ring 12 test failed<o:p></o:p></pre> <pre> [drm:amdgpu_device_ip_resume_phase2] *ERROR* resume of IP block <vce_v3_0> failed -110<o:p></o:p></pre> <pre> [drm:amdgpu_device_resume] *ERROR* amdgpu_device_ip_resume failed (-110).<o:p></o:p></pre> <pre><o:p> </o:p></pre> <pre>[How]<o:p></o:p></pre> <pre>fetch rptr appropriately, meaning move its read location further down<o:p></o:p></pre> <pre>in the code flow.<o:p></o:p></pre> <pre>With this patch applied the s3 failure is no more seen for >5k s3 cycles,<o:p></o:p></pre> <pre>which otherwise is pretty consistent.<o:p></o:p></pre> <pre><o:p> </o:p></pre> <pre>Signed-off-by: Louis Li <a href="mailto:Ching-shih.Li@amd.com" moz-do-not-send="true"><Ching-shih.Li@amd.com></a><o:p></o:p></pre> <pre>---<o:p></o:p></pre> <pre> drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c | 2 ++<o:p></o:p></pre> <pre> 1 file changed, 2 insertions(+)<o:p></o:p></pre> <pre><o:p> </o:p></pre> <pre>diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c<o:p></o:p></pre> <pre>index c021b11..92f9d46 100644<o:p></o:p></pre> <pre>--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c<o:p></o:p></pre> <pre>+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c<o:p></o:p></pre> <pre>@@ -1084,6 +1084,8 @@ int amdgpu_vce_ring_test_ring(struct amdgpu_ring *ring)<o:p></o:p></pre> <pre> if (r)<o:p></o:p></pre> <pre> return r;<o:p></o:p></pre> <pre> <o:p></o:p></pre> <pre>+ rptr = amdgpu_ring_get_rptr(ring);<o:p></o:p></pre> <pre>+<o:p></o:p></pre> </blockquote> The rptr update is there:<o:p></o:p> <pre><code> uint32_t rptr = amdgpu_ring_get_rptr(ring);<o:p></o:p></code></pre> <pre><code><o:p> </o:p></code></pre> <pre><code>Are you sure this is the root cause?<o:p></o:p></code></pre> <pre><code><o:p> </o:p></code></pre> <pre><code>Regards,<o:p></o:p></code></pre> <pre><code>Leo<o:p></o:p></code></pre> <pre><code><o:p> </o:p></code></pre> <o:p> </o:p> <blockquote style="margin-top:5.0pt;margin-bottom:5.0pt"> <pre><o:p> </o:p></pre> <pre> amdgpu_ring_write(ring, VCE_CMD_END);<o:p></o:p></pre> <pre> amdgpu_ring_commit(ring);<o:p></o:p></pre> <pre> <o:p></o:p></pre> </blockquote> </div> </blockquote> </body> </html>