From patchwork Fri Mar 23 12:57:07 2018 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Peter Ujfalusi X-Patchwork-Id: 10304553 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork.web.codeaurora.org (Postfix) with ESMTP id 7FB3A60384 for ; Fri, 23 Mar 2018 12:57:09 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 6FCF328E49 for ; Fri, 23 Mar 2018 12:57:09 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id 64B7128E53; Fri, 23 Mar 2018 12:57:09 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-4.1 required=2.0 tests=BAYES_00,DKIM_SIGNED, RCVD_IN_DNSWL_MED,T_DKIM_INVALID autolearn=ham version=3.3.1 Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher DHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.wl.linuxfoundation.org (Postfix) with ESMTPS id D7B6D28E49 for ; Fri, 23 Mar 2018 12:57:08 +0000 (UTC) Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 8CE036E371; Fri, 23 Mar 2018 12:57:07 +0000 (UTC) X-Original-To: dri-devel@lists.freedesktop.org Delivered-To: dri-devel@lists.freedesktop.org Received: from lelnx193.ext.ti.com (lelnx193.ext.ti.com [198.47.27.77]) by gabe.freedesktop.org (Postfix) with ESMTPS id 47D0A6E323 for ; Fri, 23 Mar 2018 12:57:05 +0000 (UTC) Received: from dlelxv90.itg.ti.com ([172.17.2.17]) by lelnx193.ext.ti.com (8.15.1/8.15.1) with ESMTP id w2NCv2PL030881; Fri, 23 Mar 2018 07:57:02 -0500 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=ti.com; s=ti-com-17Q1; t=1521809822; bh=BxZhmiWvPH12UN8fBzfejFAtxxVGSrwuO4mAFlSrqN4=; h=From:To:CC:Subject:Date; b=d3gVPaITyIcfZvmSHLLej3XvdaGezMSepfst35T5pDfa7ctJcMI/uAXCDh6XuiDip XEg8GeY6grIUZNIW0sAw5y8UH8Yd9o1GU0sICFchm5aZZ+gdT9R3fb05SUoAp+H132 Dp31h4etWBWJb79xtjUmAv91l59yiekDxZiY2//8= Received: from DFLE103.ent.ti.com (dfle103.ent.ti.com [10.64.6.24]) by dlelxv90.itg.ti.com (8.14.3/8.13.8) with ESMTP id w2NCv2bY003926; Fri, 23 Mar 2018 07:57:02 -0500 Received: from DFLE111.ent.ti.com (10.64.6.32) by DFLE103.ent.ti.com (10.64.6.24) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_CBC_SHA256_P256) id 15.1.1261.35; Fri, 23 Mar 2018 07:57:01 -0500 Received: from dlep33.itg.ti.com (157.170.170.75) by DFLE111.ent.ti.com (10.64.6.32) with Microsoft SMTP Server (version=TLS1_0, cipher=TLS_RSA_WITH_AES_256_CBC_SHA) id 15.1.1261.35 via Frontend Transport; Fri, 23 Mar 2018 07:57:01 -0500 Received: from feketebors.ti.com (ileax41-snat.itg.ti.com [10.172.224.153]) by dlep33.itg.ti.com (8.14.3/8.13.8) with ESMTP id w2NCuxH4026301; Fri, 23 Mar 2018 07:57:00 -0500 From: Peter Ujfalusi To: , Subject: [PATCH] drm/omap: fix memory barrier bug in DMM driver Date: Fri, 23 Mar 2018 14:57:07 +0200 Message-ID: <20180323125707.25223-1-peter.ujfalusi@ti.com> X-Mailer: git-send-email 2.16.3 MIME-Version: 1.0 X-EXCLAIMER-MD-CONFIG: e1e8a2fd-e40a-4ac6-ac9b-f7e9cc9ee180 X-BeenThere: dri-devel@lists.freedesktop.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Direct Rendering Infrastructure - Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: jsarha@ti.com, dri-devel@lists.freedesktop.org Errors-To: dri-devel-bounces@lists.freedesktop.org Sender: "dri-devel" X-Virus-Scanned: ClamAV using ClamSMTP From: Tomi Valkeinen A DMM timeout "timed out waiting for done" has been observed on DRA7 devices. The timeout happens rarely, and only when the system is under heavy load. Debugging showed that the timeout can be made to happen much more frequently by optimizing the DMM driver, so that there's almost no code between writing the last DMM descriptors to RAM, and writing to DMM register which starts the DMM transaction. The current theory is that a wmb() does not properly ensure that the data written to RAM is observable by all the components in the system. This DMM timeout has caused interesting (and rare) bugs as the error handling was not functioning properly (the error handling has been fixed in previous commits): * If a DMM timeout happened when a GEM buffer was being pinned for display on the screen, a timeout error would be shown, but the driver would continue programming DSS HW with broken buffer, leading to SYNCLOST floods and possible crashes. * If a DMM timeout happened when other user (say, video decoder) was pinning a GEM buffer, a timeout would be shown but if the user handled the error properly, no other issues followed. * If a DMM timeout happened when a GEM buffer was being released, the driver does not even notice the error, leading to crashes or hang later. This patch adds wmb() and readl() calls after the last bit is written to RAM, which should ensure that the execution proceeds only after the data is actually in RAM, and thus observable by DMM. The read-back should not be needed. Further study is required to understand if DMM is somehow special case and read-back is ok, or if DRA7's memory barriers do not work correctly. Signed-off-by: Tomi Valkeinen Signed-off-by: Peter Ujfalusi --- drivers/gpu/drm/omapdrm/omap_dmm_tiler.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c b/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c index c40f90d2db82..27c67bc36203 100644 --- a/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c +++ b/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c @@ -410,6 +410,17 @@ static int dmm_txn_commit(struct dmm_txn *txn, bool wait) } txn->last_pat->next_pa = 0; + /* ensure that the written descriptors are visible to DMM */ + wmb(); + + /* + * NOTE: the wmb() above should be enough, but there seems to be a bug + * in OMAP's memory barrier implementation, which in some rare cases may + * cause the writes not to be observable after wmb(). + */ + + /* read back to ensure the data is in RAM */ + readl(&txn->last_pat->next_pa); /* write to PAT_DESCR to clear out any pending transaction */ dmm_write(dmm, 0x0, reg[PAT_DESCR][engine->id]);