From patchwork Tue Feb 18 18:15:54 2014 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: bradley.d.volkin@intel.com X-Patchwork-Id: 3673861 Return-Path: X-Original-To: patchwork-intel-gfx@patchwork.kernel.org Delivered-To: patchwork-parsemail@patchwork2.web.kernel.org Received: from mail.kernel.org (mail.kernel.org [198.145.19.201]) by patchwork2.web.kernel.org (Postfix) with ESMTP id 2641FBF13A for ; Tue, 18 Feb 2014 18:18:28 +0000 (UTC) Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id EC3E82014A for ; Tue, 18 Feb 2014 18:18:26 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) by mail.kernel.org (Postfix) with ESMTP id C91E12010C for ; Tue, 18 Feb 2014 18:18:25 +0000 (UTC) Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 8A2AAFABCD; Tue, 18 Feb 2014 10:18:24 -0800 (PST) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from mga11.intel.com (mga11.intel.com [192.55.52.93]) by gabe.freedesktop.org (Postfix) with ESMTP id EE6EBFABC2 for ; Tue, 18 Feb 2014 10:17:32 -0800 (PST) Received: from fmsmga002.fm.intel.com ([10.253.24.26]) by fmsmga102.fm.intel.com with ESMTP; 18 Feb 2014 10:17:00 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.97,502,1389772800"; d="scan'208";a="483560853" Received: from bdvolkin-cube.ra.intel.com (HELO bdvolkin-ubuntu-desktop.ra.intel.com) ([10.10.34.187]) by fmsmga002.fm.intel.com with ESMTP; 18 Feb 2014 10:16:36 -0800 From: bradley.d.volkin@intel.com To: intel-gfx@lists.freedesktop.org Date: Tue, 18 Feb 2014 10:15:54 -0800 Message-Id: <1392747357-25703-11-git-send-email-bradley.d.volkin@intel.com> X-Mailer: git-send-email 1.8.3.2 In-Reply-To: <1392747357-25703-1-git-send-email-bradley.d.volkin@intel.com> References: <1392747357-25703-1-git-send-email-bradley.d.volkin@intel.com> Subject: [Intel-gfx] [PATCH 10/13] drm/i915: Enable PPGTT command parser checks X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.13 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Sender: intel-gfx-bounces@lists.freedesktop.org Errors-To: intel-gfx-bounces@lists.freedesktop.org X-Spam-Status: No, score=-4.8 required=5.0 tests=BAYES_00, RCVD_IN_DNSWL_MED, RP_MATCHES_RCVD, UNPARSEABLE_RELAY autolearn=unavailable version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mail.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP From: Brad Volkin Various commands that access memory have a bit to determine whether the graphics address specified in the command should use the GGTT or PPGTT for translation. These checks ensure that the bit indicates PPGTT translation. Most of these checks use the existing bit-checking infrastructure. The PIPE_CONTROL and MI_FLUSH_DW commands, however, are multi-function commands. The GGTT/PPGTT bit is only relevant for certain uses of the command. As such, this change also extends the bit-checking code to include a "condition" mask and offset. If the condition mask is non-zero then the parser only performs the bit check when the bits specified by the condition mask/offset are also non-zero. NOTE: At this point in the series PPGTT must be enabled for the parser to work correctly. If it's not enabled, userspace will not be setting the PPGTT bits the way the parser requires. VLV is the only platform where this is a problem, so at this point, we disable parsing for VLV. v2: whitespace and trailing commas fixes, rebased OTC-Tracker: AXIA-4631 Change-Id: I3f4c76b6734f1956ec47e698230f97d0998ff92b Signed-off-by: Brad Volkin --- drivers/gpu/drm/i915/i915_cmd_parser.c | 128 ++++++++++++++++++++++++++++++--- drivers/gpu/drm/i915/i915_drv.h | 6 ++ drivers/gpu/drm/i915/i915_reg.h | 6 ++ 3 files changed, 129 insertions(+), 11 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_cmd_parser.c b/drivers/gpu/drm/i915/i915_cmd_parser.c index 0351df1..1528549 100644 --- a/drivers/gpu/drm/i915/i915_cmd_parser.c +++ b/drivers/gpu/drm/i915/i915_cmd_parser.c @@ -124,10 +124,20 @@ static const struct drm_i915_cmd_descriptor common_cmds[] = { CMD( MI_STORE_DWORD_INDEX, SMI, !F, 0xFF, R ), CMD( MI_LOAD_REGISTER_IMM(1), SMI, !F, 0xFF, W, .reg = { .offset = 1, .mask = 0x007FFFFC } ), - CMD( MI_STORE_REGISTER_MEM(1), SMI, !F, 0xFF, W, - .reg = { .offset = 1, .mask = 0x007FFFFC } ), - CMD( MI_LOAD_REGISTER_MEM, SMI, !F, 0xFF, W, - .reg = { .offset = 1, .mask = 0x007FFFFC } ), + CMD( MI_STORE_REGISTER_MEM(1), SMI, !F, 0xFF, W | B, + .reg = { .offset = 1, .mask = 0x007FFFFC }, + .bits = {{ + .offset = 0, + .mask = MI_GLOBAL_GTT, + .expected = 0, + }}, ), + CMD( MI_LOAD_REGISTER_MEM, SMI, !F, 0xFF, W | B, + .reg = { .offset = 1, .mask = 0x007FFFFC }, + .bits = {{ + .offset = 0, + .mask = MI_GLOBAL_GTT, + .expected = 0, + }}, ), CMD( MI_BATCH_BUFFER_START, SMI, !F, 0xFF, S ), }; @@ -139,9 +149,31 @@ static const struct drm_i915_cmd_descriptor render_cmds[] = { CMD( MI_DISPLAY_FLIP, SMI, !F, 0xFF, R ), CMD( MI_SET_CONTEXT, SMI, !F, 0xFF, R ), CMD( MI_URB_CLEAR, SMI, !F, 0xFF, S ), + CMD( MI_STORE_DWORD_IMM, SMI, !F, 0x3F, B, + .bits = {{ + .offset = 0, + .mask = MI_GLOBAL_GTT, + .expected = 0, + }}, ), CMD( MI_UPDATE_GTT, SMI, !F, 0xFF, R ), - CMD( MI_CLFLUSH, SMI, !F, 0x3FF, S ), - CMD( MI_CONDITIONAL_BATCH_BUFFER_END, SMI, !F, 0xFF, S ), + CMD( MI_CLFLUSH, SMI, !F, 0x3FF, B, + .bits = {{ + .offset = 0, + .mask = MI_GLOBAL_GTT, + .expected = 0, + }}, ), + CMD( MI_REPORT_PERF_COUNT, SMI, !F, 0x3F, B, + .bits = {{ + .offset = 1, + .mask = MI_REPORT_PERF_COUNT_GGTT, + .expected = 0, + }}, ), + CMD( MI_CONDITIONAL_BATCH_BUFFER_END, SMI, !F, 0xFF, B, + .bits = {{ + .offset = 0, + .mask = MI_GLOBAL_GTT, + .expected = 0, + }}, ), CMD( GFX_OP_3DSTATE_VF_STATISTICS, S3D, F, 1, S ), CMD( PIPELINE_SELECT, S3D, F, 1, S ), CMD( MEDIA_VFE_STATE, S3D, !F, 0xFFFF, B, @@ -158,6 +190,13 @@ static const struct drm_i915_cmd_descriptor render_cmds[] = { .offset = 1, .mask = (PIPE_CONTROL_MMIO_WRITE | PIPE_CONTROL_NOTIFY), .expected = 0, + }, + { + .offset = 1, + .mask = PIPE_CONTROL_GLOBAL_GTT_IVB, + .expected = 0, + .condition_offset = 1, + .condition_mask = PIPE_CONTROL_POST_SYNC_OP_MASK, }}, ), }; @@ -184,15 +223,32 @@ static const struct drm_i915_cmd_descriptor hsw_render_cmds[] = { static const struct drm_i915_cmd_descriptor video_cmds[] = { CMD( MI_ARB_ON_OFF, SMI, F, 1, R ), - CMD( MI_STORE_DWORD_IMM, SMI, !F, 0xFF, S ), + CMD( MI_STORE_DWORD_IMM, SMI, !F, 0xFF, B, + .bits = {{ + .offset = 0, + .mask = MI_GLOBAL_GTT, + .expected = 0, + }}, ), CMD( MI_UPDATE_GTT, SMI, !F, 0x3F, R ), CMD( MI_FLUSH_DW, SMI, !F, 0x3F, B, .bits = {{ .offset = 0, .mask = MI_FLUSH_DW_NOTIFY, .expected = 0, + }, + { + .offset = 1, + .mask = MI_FLUSH_DW_USE_GTT, + .expected = 0, + .condition_offset = 0, + .condition_mask = MI_FLUSH_DW_OP_MASK, + }}, ), + CMD( MI_CONDITIONAL_BATCH_BUFFER_END, SMI, !F, 0xFF, B, + .bits = {{ + .offset = 0, + .mask = MI_GLOBAL_GTT, + .expected = 0, }}, ), - CMD( MI_CONDITIONAL_BATCH_BUFFER_END, SMI, !F, 0xFF, S ), /* * MFX_WAIT doesn't fit the way we handle length for most commands. * It has a length field but it uses a non-standard length bias. @@ -203,26 +259,55 @@ static const struct drm_i915_cmd_descriptor video_cmds[] = { static const struct drm_i915_cmd_descriptor vecs_cmds[] = { CMD( MI_ARB_ON_OFF, SMI, F, 1, R ), - CMD( MI_STORE_DWORD_IMM, SMI, !F, 0xFF, S ), + CMD( MI_STORE_DWORD_IMM, SMI, !F, 0xFF, B, + .bits = {{ + .offset = 0, + .mask = MI_GLOBAL_GTT, + .expected = 0, + }}, ), CMD( MI_UPDATE_GTT, SMI, !F, 0x3F, R ), CMD( MI_FLUSH_DW, SMI, !F, 0x3F, B, .bits = {{ .offset = 0, .mask = MI_FLUSH_DW_NOTIFY, .expected = 0, + }, + { + .offset = 1, + .mask = MI_FLUSH_DW_USE_GTT, + .expected = 0, + .condition_offset = 0, + .condition_mask = MI_FLUSH_DW_OP_MASK, + }}, ), + CMD( MI_CONDITIONAL_BATCH_BUFFER_END, SMI, !F, 0xFF, B, + .bits = {{ + .offset = 0, + .mask = MI_GLOBAL_GTT, + .expected = 0, }}, ), - CMD( MI_CONDITIONAL_BATCH_BUFFER_END, SMI, !F, 0xFF, S ), }; static const struct drm_i915_cmd_descriptor blt_cmds[] = { CMD( MI_DISPLAY_FLIP, SMI, !F, 0xFF, R ), - CMD( MI_STORE_DWORD_IMM, SMI, !F, 0x3FF, S ), + CMD( MI_STORE_DWORD_IMM, SMI, !F, 0x3FF, B, + .bits = {{ + .offset = 0, + .mask = MI_GLOBAL_GTT, + .expected = 0, + }}, ), CMD( MI_UPDATE_GTT, SMI, !F, 0x3F, R ), CMD( MI_FLUSH_DW, SMI, !F, 0x3F, B, .bits = {{ .offset = 0, .mask = MI_FLUSH_DW_NOTIFY, .expected = 0, + }, + { + .offset = 1, + .mask = MI_FLUSH_DW_USE_GTT, + .expected = 0, + .condition_offset = 0, + .condition_mask = MI_FLUSH_DW_OP_MASK, }}, ), CMD( COLOR_BLT, S2D, !F, 0x3F, S ), CMD( SRC_COPY_BLT, S2D, !F, 0x3F, S ), @@ -617,10 +702,21 @@ finish: */ bool i915_needs_cmd_parser(struct intel_ring_buffer *ring) { + drm_i915_private_t *dev_priv = + (drm_i915_private_t *)ring->dev->dev_private; + /* No command tables indicates a platform without parsing */ if (!ring->cmd_tables) return false; + /* + * XXX: VLV is Gen7 and therefore has cmd_tables, but has PPGTT + * disabled. That will cause all of the parser's PPGTT checks to + * fail. For now, disable parsing when PPGTT is off. + */ + if (!dev_priv->mm.aliasing_ppgtt) + return false; + return (i915.enable_cmd_parser == 1); } @@ -737,6 +833,16 @@ int i915_parse_cmds(struct intel_ring_buffer *ring, if (desc->bits[i].mask == 0) break; + if (desc->bits[i].condition_mask != 0) { + u32 offset = + desc->bits[i].condition_offset; + u32 condition = cmd[offset] & + desc->bits[i].condition_mask; + + if (condition == 0) + continue; + } + dword = cmd[desc->bits[i].offset] & desc->bits[i].mask; diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 376472f..27a48d9 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -1868,11 +1868,17 @@ struct drm_i915_cmd_descriptor { * the expected value, the parser rejects it. Only valid if flags has * the CMD_DESC_BITMASK bit set. Only entries where mask is non-zero * are valid. + * + * If the check specifies a non-zero condition_mask then the parser + * only performs the check when the bits specified by condition_mask + * are non-zero. */ struct { u32 offset; u32 mask; u32 expected; + u32 condition_offset; + u32 condition_mask; } bits[MAX_CMD_DESC_BITMASKS]; }; diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index e6dd7e9..e683b31 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -190,6 +190,8 @@ * Memory interface instructions used by the kernel */ #define MI_INSTR(opcode, flags) (((opcode) << 23) | (flags)) +/* Many MI commands use bit 22 of the header dword for GGTT vs PPGTT */ +#define MI_GLOBAL_GTT (1<<22) #define MI_NOOP MI_INSTR(0, 0) #define MI_USER_INTERRUPT MI_INSTR(0x02, 0) @@ -269,6 +271,7 @@ #define MI_FLUSH_DW_STORE_INDEX (1<<21) #define MI_INVALIDATE_TLB (1<<18) #define MI_FLUSH_DW_OP_STOREDW (1<<14) +#define MI_FLUSH_DW_OP_MASK (3<<14) #define MI_FLUSH_DW_NOTIFY (1<<8) #define MI_INVALIDATE_BSD (1<<7) #define MI_FLUSH_DW_USE_GTT (1<<2) @@ -335,6 +338,7 @@ #define PIPE_CONTROL_CS_STALL (1<<20) #define PIPE_CONTROL_TLB_INVALIDATE (1<<18) #define PIPE_CONTROL_QW_WRITE (1<<14) +#define PIPE_CONTROL_POST_SYNC_OP_MASK (3<<14) #define PIPE_CONTROL_DEPTH_STALL (1<<13) #define PIPE_CONTROL_WRITE_FLUSH (1<<12) #define PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH (1<<12) /* gen6+ */ @@ -363,6 +367,8 @@ #define MI_URB_CLEAR MI_INSTR(0x19, 0) #define MI_UPDATE_GTT MI_INSTR(0x23, 0) #define MI_CLFLUSH MI_INSTR(0x27, 0) +#define MI_REPORT_PERF_COUNT MI_INSTR(0x28, 0) +#define MI_REPORT_PERF_COUNT_GGTT (1<<0) #define MI_LOAD_REGISTER_MEM MI_INSTR(0x29, 0) #define MI_LOAD_REGISTER_REG MI_INSTR(0x2A, 0) #define MI_RS_STORE_DATA_IMM MI_INSTR(0x2B, 0)