From: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com> To: "Yang, Fei" <fei.yang@intel.com>, "Roper, Matthew D" <matthew.d.roper@intel.com> Cc: "Morek, Szymon" <szymon.morek@intel.com>, "intel-gfx@lists.freedesktop.org" <intel-gfx@lists.freedesktop.org>, "dri-devel@lists.freedesktop.org" <dri-devel@lists.freedesktop.org>, "Ceraolo Spurio, Daniele" <daniele.ceraolospurio@intel.com>, "Justen, Jordan L" <jordan.l.justen@intel.com>, "Belgaumkar, Vinay" <vinay.belgaumkar@intel.com>, "Kumar Valsan, Prathap" <prathap.kumar.valsan@intel.com> Subject: Re: [PATCH 1/2] drm/i915/xehp: Add compute engine ABI Date: Tue, 26 Apr 2022 08:25:10 +0100 [thread overview] Message-ID: <d13dbabb-0848-7f60-ef01-f1d2a8c77da4@linux.intel.com> (raw) In-Reply-To: <BYAPR11MB2567CB015A6B1ADF5F1FEDAB9AF89@BYAPR11MB2567.namprd11.prod.outlook.com> On 25/04/2022 19:40, Yang, Fei wrote: >>>> --- a/drivers/gpu/drm/i915/gt/intel_gt.c >>>> +++ b/drivers/gpu/drm/i915/gt/intel_gt.c >>>> @@ -1175,6 +1175,7 @@ void intel_gt_invalidate_tlbs(struct intel_gt *gt) >>>> [VIDEO_DECODE_CLASS] = GEN12_VD_TLB_INV_CR, >>>> [VIDEO_ENHANCEMENT_CLASS] = GEN12_VE_TLB_INV_CR, >>>> [COPY_ENGINE_CLASS] = GEN12_BLT_TLB_INV_CR, >>>> + [COMPUTE_CLASS] = GEN12_GFX_TLB_INV_CR, >>> >>> Do you know what 0xcf04 is? > > Looks like that is the TLB invalidation register for each compute context. What does compute "context" stand for in this context, as used in bspec? Not compute command streamer? Suspiciously individual bits (eight of them) are reserved per context, just like for example in GEN12_VD_TLB_INV_CR. >>> Or if GEN12_GFX_TLB_INV_CR is correct then I think get_reg_and_bit() >>> might need adjusting to always select bit 0 for any compute engine >>> instance. Not sure how hardware would behave if value other than '1' >>> would be written into 0xced8. >> >> I think Prathap and Fei have more familiarity with the MMIO TLB invalidation; adding them for their thoughts. > > I believe GEN12_GFX_TLB_INV_CR is the right one to use because we are invalidating the TLB for each engine. I don't understand this argument, I guess because I don't understand 0xcf04 still. > I'm not sure if we could narrow down to exact which compute context the TLB needs to be invalidated though. If that's possible it might be a bit more efficient. Or even correct if 0xcf04 is for compute command streamers? That's my concern. Regards, Tvrtko
WARNING: multiple messages have this Message-ID (diff)
From: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com> To: "Yang, Fei" <fei.yang@intel.com>, "Roper, Matthew D" <matthew.d.roper@intel.com> Cc: "Morek, Szymon" <szymon.morek@intel.com>, "intel-gfx@lists.freedesktop.org" <intel-gfx@lists.freedesktop.org>, "dri-devel@lists.freedesktop.org" <dri-devel@lists.freedesktop.org> Subject: Re: [Intel-gfx] [PATCH 1/2] drm/i915/xehp: Add compute engine ABI Date: Tue, 26 Apr 2022 08:25:10 +0100 [thread overview] Message-ID: <d13dbabb-0848-7f60-ef01-f1d2a8c77da4@linux.intel.com> (raw) In-Reply-To: <BYAPR11MB2567CB015A6B1ADF5F1FEDAB9AF89@BYAPR11MB2567.namprd11.prod.outlook.com> On 25/04/2022 19:40, Yang, Fei wrote: >>>> --- a/drivers/gpu/drm/i915/gt/intel_gt.c >>>> +++ b/drivers/gpu/drm/i915/gt/intel_gt.c >>>> @@ -1175,6 +1175,7 @@ void intel_gt_invalidate_tlbs(struct intel_gt *gt) >>>> [VIDEO_DECODE_CLASS] = GEN12_VD_TLB_INV_CR, >>>> [VIDEO_ENHANCEMENT_CLASS] = GEN12_VE_TLB_INV_CR, >>>> [COPY_ENGINE_CLASS] = GEN12_BLT_TLB_INV_CR, >>>> + [COMPUTE_CLASS] = GEN12_GFX_TLB_INV_CR, >>> >>> Do you know what 0xcf04 is? > > Looks like that is the TLB invalidation register for each compute context. What does compute "context" stand for in this context, as used in bspec? Not compute command streamer? Suspiciously individual bits (eight of them) are reserved per context, just like for example in GEN12_VD_TLB_INV_CR. >>> Or if GEN12_GFX_TLB_INV_CR is correct then I think get_reg_and_bit() >>> might need adjusting to always select bit 0 for any compute engine >>> instance. Not sure how hardware would behave if value other than '1' >>> would be written into 0xced8. >> >> I think Prathap and Fei have more familiarity with the MMIO TLB invalidation; adding them for their thoughts. > > I believe GEN12_GFX_TLB_INV_CR is the right one to use because we are invalidating the TLB for each engine. I don't understand this argument, I guess because I don't understand 0xcf04 still. > I'm not sure if we could narrow down to exact which compute context the TLB needs to be invalidated though. If that's possible it might be a bit more efficient. Or even correct if 0xcf04 is for compute command streamers? That's my concern. Regards, Tvrtko
next prev parent reply other threads:[~2022-04-26 7:25 UTC|newest] Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top 2022-04-22 19:50 [PATCH 0/2] i915: Turn on compute engine support Matt Roper 2022-04-22 19:50 ` [Intel-gfx] " Matt Roper 2022-04-22 19:50 ` [PATCH 1/2] drm/i915/xehp: Add compute engine ABI Matt Roper 2022-04-22 19:50 ` [Intel-gfx] " Matt Roper 2022-04-25 10:41 ` Tvrtko Ursulin 2022-04-25 10:41 ` [Intel-gfx] " Tvrtko Ursulin 2022-04-25 17:35 ` Matt Roper 2022-04-25 17:35 ` [Intel-gfx] " Matt Roper 2022-04-25 18:40 ` Yang, Fei 2022-04-25 18:40 ` [Intel-gfx] " Yang, Fei 2022-04-26 7:25 ` Tvrtko Ursulin [this message] 2022-04-26 7:25 ` Tvrtko Ursulin 2022-04-28 0:27 ` Kumar Valsan, Prathap 2022-04-28 3:44 ` Matt Roper 2022-04-28 3:44 ` [Intel-gfx] " Matt Roper 2022-04-25 14:48 ` Andi Shyti 2022-04-22 19:50 ` [PATCH 2/2] drm/i915: Xe_HP SDV and DG2 have up to 4 CCS engines Matt Roper 2022-04-22 19:50 ` [Intel-gfx] " Matt Roper 2022-04-25 14:48 ` Andi Shyti 2022-04-22 20:20 ` [Intel-gfx] ✗ Fi.CI.SPARSE: warning for i915: Turn on compute engine support Patchwork 2022-04-22 20:47 ` [Intel-gfx] ✗ Fi.CI.BAT: failure " Patchwork 2022-04-22 21:32 ` [Intel-gfx] ✗ Fi.CI.SPARSE: warning for i915: Turn on compute engine support (rev2) Patchwork 2022-04-22 22:07 ` [Intel-gfx] ✗ Fi.CI.BAT: failure " Patchwork 2022-04-22 22:33 ` Matt Roper 2022-04-22 23:02 ` [Intel-gfx] ✗ Fi.CI.SPARSE: warning for i915: Turn on compute engine support (rev3) Patchwork 2022-04-22 23:24 ` [Intel-gfx] ✓ Fi.CI.BAT: success " Patchwork 2022-04-23 1:18 ` [Intel-gfx] ✗ Fi.CI.IGT: failure " Patchwork
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=d13dbabb-0848-7f60-ef01-f1d2a8c77da4@linux.intel.com \ --to=tvrtko.ursulin@linux.intel.com \ --cc=daniele.ceraolospurio@intel.com \ --cc=dri-devel@lists.freedesktop.org \ --cc=fei.yang@intel.com \ --cc=intel-gfx@lists.freedesktop.org \ --cc=jordan.l.justen@intel.com \ --cc=matthew.d.roper@intel.com \ --cc=prathap.kumar.valsan@intel.com \ --cc=szymon.morek@intel.com \ --cc=vinay.belgaumkar@intel.com \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.