Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 370ACC6FD1C for ; Tue, 14 Mar 2023 13:03:21 +0000 (UTC) Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 0F11910E7DE; Tue, 14 Mar 2023 13:03:21 +0000 (UTC) Received: from mga03.intel.com (mga03.intel.com [134.134.136.65]) by gabe.freedesktop.org (Postfix) with ESMTPS id E573510E7DC for ; Tue, 14 Mar 2023 13:03:18 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1678798998; x=1710334998; h=message-id:date:mime-version:subject:to:cc:references: from:in-reply-to:content-transfer-encoding; bh=Lg3SPoDA4fNlAnGpqj9I6x5fJn+eGmNPE+o9QUxCrhk=; b=dNT8PT7k+LqaTY99cIQkgx/OCQlmXj4s5UDnJjRpMT1un+zk5XI+FT1I pHMDO/QV49pHI5nCwDmUsaVb1Gb/4Xlyq8Pn2BhT9QN3/rR/GctJG2bBS L/T5HXI2LaDjcOyCWxFV5e3V6xW2ncuq+xbuE8kiTUb1je/SvafU8k5bb c+L75pe02L0saHZqdANynfShkw5ZHxbj7/t8tTmKFmmDPFUC0YvaCIAfq hpibO6aj01v4xUHiBGcXwr4nSUWT7ECrRnMAkE6zNmvfEVyDlnS5firlM iYdMVptNH6X8x0vM6hAInerLElxX8f9y2kKLAgIiPrh13O6MP/9jIkICF A==; X-IronPort-AV: E=McAfee;i="6500,9779,10648"; a="339780062" X-IronPort-AV: E=Sophos;i="5.98,259,1673942400"; d="scan'208";a="339780062" Received: from fmsmga003.fm.intel.com ([10.253.24.29]) by orsmga103.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 14 Mar 2023 06:03:18 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6500,9779,10648"; a="768079077" X-IronPort-AV: E=Sophos;i="5.98,259,1673942400"; d="scan'208";a="768079077" Received: from fathuddx-mobl.gar.corp.intel.com (HELO [10.249.254.202]) ([10.249.254.202]) by fmsmga003-auth.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 14 Mar 2023 06:03:17 -0700 Message-ID: Date: Tue, 14 Mar 2023 14:03:14 +0100 MIME-Version: 1.0 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101 Thunderbird/102.8.0 Content-Language: en-US To: Matthew Auld References: <20230313204519.158888-1-thomas.hellstrom@linux.intel.com> <20230313204519.158888-2-thomas.hellstrom@linux.intel.com> From: =?UTF-8?Q?Thomas_Hellstr=c3=b6m?= In-Reply-To: Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 8bit Subject: Re: [Intel-xe] [PATCH 1/6] drm/xe: Use a define to set initial seqno for fences X-BeenThere: intel-xe@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel Xe graphics driver List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: intel-xe@lists.freedesktop.org Errors-To: intel-xe-bounces@lists.freedesktop.org Sender: "Intel-xe" On 3/14/23 12:54, Matthew Auld wrote: > On Mon, 13 Mar 2023 at 20:46, Thomas Hellström > wrote: >> Also for HW fences, write the initial seqno - 1 to the HW completed >> seqno to initialize. >> >> Signed-off-by: Thomas Hellström >> --- >> drivers/gpu/drm/xe/xe_engine.c | 5 +++-- >> drivers/gpu/drm/xe/xe_hw_fence.c | 5 ++--- >> drivers/gpu/drm/xe/xe_hw_fence.h | 2 ++ >> drivers/gpu/drm/xe/xe_lrc.c | 3 +++ >> 4 files changed, 10 insertions(+), 5 deletions(-) >> >> diff --git a/drivers/gpu/drm/xe/xe_engine.c b/drivers/gpu/drm/xe/xe_engine.c >> index 3b6f8a25112a..7e6367a84f1f 100644 >> --- a/drivers/gpu/drm/xe/xe_engine.c >> +++ b/drivers/gpu/drm/xe/xe_engine.c >> @@ -13,6 +13,7 @@ >> >> #include "xe_device.h" >> #include "xe_gt.h" >> +#include "xe_hw_fence.h" >> #include "xe_lrc.h" >> #include "xe_macros.h" >> #include "xe_migrate.h" >> @@ -57,11 +58,11 @@ static struct xe_engine *__xe_engine_create(struct xe_device *xe, >> >> if (xe_engine_is_parallel(e)) { >> e->parallel.composite_fence_ctx = dma_fence_context_alloc(1); >> - e->parallel.composite_fence_seqno = 1; >> + e->parallel.composite_fence_seqno = XE_FENCE_INITIAL_SEQNO; >> } >> if (e->flags & ENGINE_FLAG_VM) { >> e->bind.fence_ctx = dma_fence_context_alloc(1); >> - e->bind.fence_seqno = 1; >> + e->bind.fence_seqno = XE_FENCE_INITIAL_SEQNO; >> } >> >> for (i = 0; i < width; ++i) { >> diff --git a/drivers/gpu/drm/xe/xe_hw_fence.c b/drivers/gpu/drm/xe/xe_hw_fence.c >> index e56ca2867545..ca8304650713 100644 >> --- a/drivers/gpu/drm/xe/xe_hw_fence.c >> +++ b/drivers/gpu/drm/xe/xe_hw_fence.c >> @@ -129,7 +129,7 @@ void xe_hw_fence_ctx_init(struct xe_hw_fence_ctx *ctx, struct xe_gt *gt, >> ctx->gt = gt; >> ctx->irq = irq; >> ctx->dma_fence_ctx = dma_fence_context_alloc(1); >> - ctx->next_seqno = 1; >> + ctx->next_seqno = XE_FENCE_INITIAL_SEQNO; >> sprintf(ctx->name, "%s", name); >> } >> >> @@ -164,8 +164,7 @@ static bool xe_hw_fence_signaled(struct dma_fence *dma_fence) >> struct xe_device *xe = gt_to_xe(fence->ctx->gt); >> u32 seqno = xe_map_rd(xe, &fence->seqno_map, 0, u32); >> >> - return dma_fence->error || >> - (s32)fence->dma.seqno <= (s32)seqno; >> + return dma_fence->error || (s32)(u32)fence->dma.seqno <= (s32)seqno; > Why do we need this change? Also the hw fence dma.seqno should always > be u32, right? Thanks for reviewing, Matt. Yes the value should always be u32, but the type is u64, and I read somewhere that casting an u64 to s32 could yield implementation-dependent results. I could double-check that, or actually even use __dma_fence_is_later(). That seems to be the best choice. /Thomas > >> } >> >> static bool xe_hw_fence_enable_signaling(struct dma_fence *dma_fence) >> diff --git a/drivers/gpu/drm/xe/xe_hw_fence.h b/drivers/gpu/drm/xe/xe_hw_fence.h >> index 07f202db6526..523c2611ef5d 100644 >> --- a/drivers/gpu/drm/xe/xe_hw_fence.h >> +++ b/drivers/gpu/drm/xe/xe_hw_fence.h >> @@ -8,6 +8,8 @@ >> >> #include "xe_hw_fence_types.h" >> >> +#define XE_FENCE_INITIAL_SEQNO 1 >> + >> int xe_hw_fence_module_init(void); >> void xe_hw_fence_module_exit(void); >> >> diff --git a/drivers/gpu/drm/xe/xe_lrc.c b/drivers/gpu/drm/xe/xe_lrc.c >> index 9140b057a5ba..fb8c6f7d6528 100644 >> --- a/drivers/gpu/drm/xe/xe_lrc.c >> +++ b/drivers/gpu/drm/xe/xe_lrc.c >> @@ -697,6 +697,9 @@ int xe_lrc_init(struct xe_lrc *lrc, struct xe_hw_engine *hwe, >> arb_enable = MI_ARB_ON_OFF | MI_ARB_ENABLE; >> xe_lrc_write_ring(lrc, &arb_enable, sizeof(arb_enable)); >> >> + map = __xe_lrc_seqno_map(lrc); >> + xe_map_write32(lrc_to_xe(lrc), &map, lrc->fence_ctx.next_seqno - 1); >> + >> return 0; >> >> err_lrc_finish: >> -- >> 2.39.2 >>