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 EB479C76196 for ; Fri, 24 Mar 2023 21:04:31 +0000 (UTC) Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id B805A10E56B; Fri, 24 Mar 2023 21:04:31 +0000 (UTC) Received: from mga09.intel.com (mga09.intel.com [134.134.136.24]) by gabe.freedesktop.org (Postfix) with ESMTPS id 5529310E1C8 for ; Fri, 24 Mar 2023 21:04:25 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1679691865; x=1711227865; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=kCBAlUcQwBPZ+io7wuiB4vjVQCAF3W1PHep2L9XruhM=; b=jF9Mrj6I+KCt+bbpIhbt6Xz3mVIcNtiLnyJ6LteVnXfo+SLazLnIB0Pk vJmGH4kRCR68gE7clUqPJp4MlbNFJ0fm/EJYvqO8zC9u+Otx+XZy6dCbS 6iRVEcN1ekIV1ToA7Q+Z4MrtTQipyjgFSoioyUHNs7+DQDU5cSphm3GGG EDO8KxX4p31zPDoOtR8eEqTQMoDcFMZ85CwTVcP4ZN/M8Fn8xoY9FqBzm p0xBVw2kf5ezjOJWp1T4kt6CS5Ik5VDw/1g814H3qVsukVYr8s5C3H/hw XYGaeAZf+bO9ZvozTKxXSRGm5S2/7Cj8qZdaNDkmqAZ8JL1NB1m1C+dom w==; X-IronPort-AV: E=McAfee;i="6600,9927,10659"; a="341456107" X-IronPort-AV: E=Sophos;i="5.98,288,1673942400"; d="scan'208";a="341456107" Received: from fmsmga004.fm.intel.com ([10.253.24.48]) by orsmga102.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 24 Mar 2023 14:04:24 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10659"; a="752022849" X-IronPort-AV: E=Sophos;i="5.98,288,1673942400"; d="scan'208";a="752022849" Received: from mdroper-desk1.fm.intel.com ([10.1.27.134]) by fmsmga004-auth.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 24 Mar 2023 14:04:24 -0700 From: Matt Roper To: intel-xe@lists.freedesktop.org Date: Fri, 24 Mar 2023 14:04:12 -0700 Message-Id: <20230324210415.2434992-4-matthew.d.roper@intel.com> X-Mailer: git-send-email 2.39.2 In-Reply-To: <20230324210415.2434992-1-matthew.d.roper@intel.com> References: <20230324210415.2434992-1-matthew.d.roper@intel.com> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Subject: [Intel-xe] [PATCH 3/6] drm/xe/pat: Handle unicast vs MCR PAT registers 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: matthew.d.roper@intel.com Errors-To: intel-xe-bounces@lists.freedesktop.org Sender: "Intel-xe" The PAT_INDEX registers are MCR registers on some platforms and unicast on others. On MTL the handling even varies between GTs: the primary GT uses MCR registers while the media GT uses unicast registers. Let's add proper MCR programming on the relevant platforms/GTs. Given that we PAT tables to change pretty regularly on future platforms, we'll make PAT programming an exception to the usual model of assuming new platforms should inherit the previous platform's behavior. Instead we'll raise a warning if the current platform isn't handled in the if/else ladder. This should help prevent subtle cache misbehavior if we forget to add the table for a new platform. Bspec: 66534, 67609, 67788 Signed-off-by: Matt Roper --- drivers/gpu/drm/xe/xe_pat.c | 35 ++++++++++++++++++++++++++++------- 1 file changed, 28 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_pat.c b/drivers/gpu/drm/xe/xe_pat.c index 75d25e0c0b9a..f511d9a54358 100644 --- a/drivers/gpu/drm/xe/xe_pat.c +++ b/drivers/gpu/drm/xe/xe_pat.c @@ -5,9 +5,10 @@ #include "regs/xe_reg_defs.h" #include "xe_gt.h" +#include "xe_gt_mcr.h" #include "xe_mmio.h" -#define GEN12_PAT_INDEX(index) _MMIO(0x4800 + (index) * 4) +#define _PAT_INDEX(index) (0x4800 + (index) * 4) #define GEN8_PPAT_WB (3<<0) #define GEN8_PPAT_WT (2<<0) @@ -56,17 +57,37 @@ const u32 mtl_pat_table[] = { #define PROGRAM_PAT_UNICAST(gt, table) do { \ for (int i = 0; i < ARRAY_SIZE(table); i++) \ - xe_mmio_write32(gt, GEN12_PAT_INDEX(i).reg, table[i]); \ + xe_mmio_write32(gt, _PAT_INDEX(i), table[i]); \ +} while (0) + +#define PROGRAM_PAT_MCR(gt, table) do { \ + for (int i = 0; i < ARRAY_SIZE(table); i++) \ + xe_gt_mcr_multicast_write(gt, MCR_REG(_PAT_INDEX(i)), table[i]); \ } while (0) void xe_pat_init(struct xe_gt *gt) { struct xe_device *xe = gt_to_xe(gt); - if (xe->info.platform == XE_METEORLAKE) - PROGRAM_PAT_UNICAST(gt, mtl_pat_table); - else if (xe->info.platform == XE_PVC) - PROGRAM_PAT_UNICAST(gt, pvc_pat_table); - else + if (xe->info.platform == XE_METEORLAKE) { + if (xe_gt_is_media_type(gt)) + PROGRAM_PAT_UNICAST(gt, mtl_pat_table); + else + PROGRAM_PAT_MCR(gt, mtl_pat_table); + } else if (xe->info.platform == XE_PVC) { + PROGRAM_PAT_MCR(gt, pvc_pat_table); + } else if (xe->info.platform == XE_DG2) { + PROGRAM_PAT_MCR(gt, pvc_pat_table); + } else if (GRAPHICS_VERx100(xe) <= 1210) { PROGRAM_PAT_UNICAST(gt, tgl_pat_table); + } else { + /* + * Going forward we expect to need new PAT settings for most + * new platforms; failure to provide a new table can easily + * lead to subtle, hard-to-debug problems. If none of the + * conditions above match the platform we're running on we'll + * raise an error rather than trying to silently inherit the + * most recent platform's behavior. + */ + } } -- 2.39.2