{"id":2219432,"url":"http://patchwork.ozlabs.org/api/patches/2219432/?format=json","web_url":"http://patchwork.ozlabs.org/project/linuxppc-dev/patch/20260402174925.v3.5.If62b84471ef2c85e7ad250f0468867d6dba965ab@changeid/","project":{"id":2,"url":"http://patchwork.ozlabs.org/api/projects/2/?format=json","name":"Linux PPC development","link_name":"linuxppc-dev","list_id":"linuxppc-dev.lists.ozlabs.org","list_email":"linuxppc-dev@lists.ozlabs.org","web_url":"https://github.com/linuxppc/wiki/wiki","scm_url":"https://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux.git","webscm_url":"https://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux.git/","list_archive_url":"https://lore.kernel.org/linuxppc-dev/","list_archive_url_format":"https://lore.kernel.org/linuxppc-dev/{}/","commit_url_format":"https://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux.git/commit/?id={}"},"msgid":"<20260402174925.v3.5.If62b84471ef2c85e7ad250f0468867d6dba965ab@changeid>","list_archive_url":"https://lore.kernel.org/linuxppc-dev/20260402174925.v3.5.If62b84471ef2c85e7ad250f0468867d6dba965ab@changeid/","date":"2026-04-03T00:49:51","name":"[v3,5/9] driver core: Replace dev->dma_ops_bypass with DEV_FLAG_DMA_OPS_BYPASS","commit_ref":null,"pull_url":null,"state":"new","archived":false,"hash":"91b37f8a1cdfa5e7b46ac072de6c23f6775e131e","submitter":{"id":9763,"url":"http://patchwork.ozlabs.org/api/people/9763/?format=json","name":"Douglas Anderson","email":"dianders@chromium.org"},"delegate":null,"mbox":"http://patchwork.ozlabs.org/project/linuxppc-dev/patch/20260402174925.v3.5.If62b84471ef2c85e7ad250f0468867d6dba965ab@changeid/mbox/","series":[{"id":498570,"url":"http://patchwork.ozlabs.org/api/series/498570/?format=json","web_url":"http://patchwork.ozlabs.org/project/linuxppc-dev/list/?series=498570","date":"2026-04-03T00:49:46","name":"driver core: Fix some race conditions","version":3,"mbox":"http://patchwork.ozlabs.org/series/498570/mbox/"}],"comments":"http://patchwork.ozlabs.org/api/patches/2219432/comments/","check":"pending","checks":"http://patchwork.ozlabs.org/api/patches/2219432/checks/","tags":{},"related":[],"headers":{"Return-Path":"\n ","X-Original-To":["incoming@patchwork.ozlabs.org","linuxppc-dev@lists.ozlabs.org"],"Delivered-To":"patchwork-incoming@legolas.ozlabs.org","Authentication-Results":["legolas.ozlabs.org;\n\tdkim=pass (1024-bit key;\n unprotected) header.d=chromium.org header.i=@chromium.org header.a=rsa-sha256\n header.s=google header.b=cykLDNip;\n\tdkim-atps=neutral","legolas.ozlabs.org;\n spf=pass (sender SPF authorized) smtp.mailfrom=lists.ozlabs.org\n (client-ip=2404:9400:21b9:f100::1; helo=lists.ozlabs.org;\n envelope-from=linuxppc-dev+bounces-19245-incoming=patchwork.ozlabs.org@lists.ozlabs.org;\n receiver=patchwork.ozlabs.org)","lists.ozlabs.org;\n arc=none smtp.remote-ip=\"2607:f8b0:4864:20::132e\"","lists.ozlabs.org;\n dmarc=pass (p=none dis=none) header.from=chromium.org","lists.ozlabs.org;\n\tdkim=pass (1024-bit key;\n unprotected) header.d=chromium.org header.i=@chromium.org header.a=rsa-sha256\n header.s=google header.b=cykLDNip;\n\tdkim-atps=neutral","lists.ozlabs.org;\n spf=pass (sender SPF authorized) smtp.mailfrom=chromium.org\n (client-ip=2607:f8b0:4864:20::132e; helo=mail-dy1-x132e.google.com;\n envelope-from=dianders@chromium.org; receiver=lists.ozlabs.org)"],"Received":["from lists.ozlabs.org (lists.ozlabs.org\n [IPv6:2404:9400:21b9:f100::1])\n\t(using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits)\n\t key-exchange x25519)\n\t(No client certificate requested)\n\tby legolas.ozlabs.org (Postfix) with ESMTPS id 4fn0Y55NGHz1yDH\n\tfor ; Fri, 03 Apr 2026 11:51:53 +1100 (AEDT)","from boromir.ozlabs.org (localhost [127.0.0.1])\n\tby lists.ozlabs.org (Postfix) with ESMTP id 4fn0Y53zkyz2yVv;\n\tFri, 03 Apr 2026 11:51:53 +1100 (AEDT)","from mail-dy1-x132e.google.com (mail-dy1-x132e.google.com\n [IPv6:2607:f8b0:4864:20::132e])\n\t(using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits)\n\t key-exchange x25519 server-signature RSA-PSS (2048 bits) server-digest\n SHA256)\n\t(No client certificate requested)\n\tby lists.ozlabs.org (Postfix) with ESMTPS id 4fn0Y465CVz2xQD\n\tfor ; Fri, 03 Apr 2026 11:51:52 +1100 (AEDT)","by mail-dy1-x132e.google.com with SMTP id\n 5a478bee46e88-2c1632faeb9so2914274eec.0\n for ;\n Thu, 02 Apr 2026 17:51:52 -0700 (PDT)","from dianders.sjc.corp.google.com\n ([2a00:79e0:2e7c:8:5db3:7542:a530:f43a])\n by smtp.gmail.com with ESMTPSA id\n 5a478bee46e88-2ca78df3b84sm3630074eec.5.2026.04.02.17.51.47\n (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256);\n Thu, 02 Apr 2026 17:51:49 -0700 (PDT)"],"ARC-Seal":"i=1; a=rsa-sha256; d=lists.ozlabs.org; s=201707; t=1775177513;\n\tcv=none;\n b=ZjdN5mFTgZ/EMPy1eP7vlu7LNx9ZMygUluZIFNXRYocOue/GgLCCm49eTQkHIjLtXVB+9O18VThxzv/isWcydIivZs+XXYCOGlcmAjcgZji6OoOM8evkRBL03m7TMjX4FxZxxK/vjlt983YNWb7mHhH8UxUZYQ29l89n1fKLGUHP8ZbTtqLASHKSQayI9k5xv3T5NELHVMt6GxZBPXbCCRmGi7vocePjDpkyyCRvUvTYJfZc5y5sj+Q3fWZknJ31j5hPKU7/sedwRrVcSS2iyQ3jhf41P124bbLEzfvgXs1vt5Z6fCO1ibD5fn0i3U9HoO8p5xc0/lNSfc1lcXeZ/A==","ARC-Message-Signature":"i=1; a=rsa-sha256; d=lists.ozlabs.org; s=201707;\n\tt=1775177513; c=relaxed/relaxed;\n\tbh=lZiTiKY0xg6ysj9D1M7VJbCxfmHcK2OdGCIMJgNJnE8=;\n\th=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References:\n\t MIME-Version;\n b=lo0phTY+UH8pAqpTsd11hBDJK5wvJS7fSxBt4x2E/lHc4/elKk9hBz1X0F4JIHthn9EwSaJkEgHdU4icleaBsKiZuxNNUVgaGQzYSKDsLVsCnLcZVzx5soECBKpkvBJCgpzH2GAVxFZb7Fd5RQQ4fGK3QSriMX1/+h46ihJbwa8YzqIGxIWQS1BnbeH4YxhAPREzDkt29bWEnRigThs20jMZyiTm2NhsU0KDivgi9VmcHYumzCA6AfKXEO0h/s/RI8cPhCA+I6S2o08pKczLG7+KcS6Zy4jjNBnXPvCYtpN/WlNS/Rm2GNHQeO2ndtFhVZ49u84pxmc0J9svZ/aszw==","ARC-Authentication-Results":"i=1; lists.ozlabs.org;\n dmarc=pass (p=none dis=none) header.from=chromium.org;\n dkim=pass (1024-bit key;\n unprotected) header.d=chromium.org header.i=@chromium.org header.a=rsa-sha256\n header.s=google header.b=cykLDNip; dkim-atps=neutral;\n spf=pass (client-ip=2607:f8b0:4864:20::132e; helo=mail-dy1-x132e.google.com;\n envelope-from=dianders@chromium.org;\n receiver=lists.ozlabs.org) smtp.mailfrom=chromium.org","DKIM-Signature":"v=1; a=rsa-sha256; c=relaxed/relaxed;\n d=chromium.org; s=google; t=1775177511; x=1775782311;\n darn=lists.ozlabs.org;\n h=content-transfer-encoding:mime-version:references:in-reply-to\n :message-id:date:subject:cc:to:from:from:to:cc:subject:date\n :message-id:reply-to;\n bh=lZiTiKY0xg6ysj9D1M7VJbCxfmHcK2OdGCIMJgNJnE8=;\n b=cykLDNipXra21PihGUd7qXWJ74WDXmGyKFqnpajIiiCkiLM7StJqQWo5pG3fcp02Rb\n yRCIt8kJRda5UpvtSXToa8Na1+4DJ9TtFjak1DlZNuTzZfv2M6WIZNbqYFAG7HSRT+Ma\n S3EiMzMujt2Iuib7YMw78ZambeWRlV3EEFLDs=","X-Google-DKIM-Signature":"v=1; a=rsa-sha256; c=relaxed/relaxed;\n d=1e100.net; s=20251104; t=1775177511; x=1775782311;\n h=content-transfer-encoding:mime-version:references:in-reply-to\n :message-id:date:subject:cc:to:from:x-gm-gg:x-gm-message-state:from\n :to:cc:subject:date:message-id:reply-to;\n bh=lZiTiKY0xg6ysj9D1M7VJbCxfmHcK2OdGCIMJgNJnE8=;\n b=U0hFHC1XeEMANMlNc8e34q6jH+VSFB5Lic3sRkEAcaVYAp7d70rawN2cnF1gCJOqx/\n rcEbFk2VvsKQM8EMJ9NjvRIh3Kai/dwz9hwuw6/l7JhaYVSFO1XfDlA1xBvDjvaHsWZI\n nzluUvrF1qcFmaWFZZxDpk4tvVjAYcsg+Z5ntpzGeremv42fL14gwOkhq0aqn8lcACjx\n biZo1Kjjkilmp+Q9oO7DKcp97PLUTIg6MRXqBCvuRxizR5JC43e+liHAyK9NddAGBrD2\n ZN9aqpGjHiaZuDzSYD0vnkFELmRvpY3bPlX1yVjQN94Ue8e1DiCEIDPTFOfxZUKb32VW\n bndw==","X-Forwarded-Encrypted":"i=1;\n AJvYcCVZL0WKWWAs20v0rkvrY6hCOCjd8u8sPwDze9WD4u5QMk48O3ud9dVX4uAaAZXaTTeJip/1ZAu34dyo51c=@lists.ozlabs.org","X-Gm-Message-State":"AOJu0YxHt3OwIoKXZ3yA95JUhRsw9clwmQ6RRMJ4gbtMw38tt77sV729\n\tfZbNuXf8HVdx0Q5R0lm6Tq0nstDIoG092039uCiW7hsx2uFxAaYzCodpQ6L408VIMQ==","X-Gm-Gg":"AeBDieuhTXhPyOY1r/SdpNq/whpuDidifJLc4rIy6uNt/u0PRZkm7kX0nyMmam5yfmr\n\tN2fWpwQrQcovIlCqgprZVtFmCBb03l7WW7trao05GDLkCdO4sDcl7gx77b5etuou57E1p+RlusE\n\tvE/NIZGxK78W9q5nPtBUf3WsYalqk6fmx+C9Zq8Emesx45sO2eOD/k1p2dc44N+IDJgckWCBIo2\n\tgG8lLuKrenSQrqq/5kvs4IrA/xLw6SiWK0MYmZeH0P9peNlPa9Y+SLtRdNZsfaP+VnjX9K+WRWJ\n\t+LLKvj0N/INlpGp753nKsgzvde4ycL57vfRWVy5a8nJlLVp3upCGdhP0lO+LQc6npyPyQVYiedu\n\ttZSm1JKCv93ziSgGYbGImaclJ6f/WOFFA119Pm/ZJObOkPbqnZpA5wStEd+i89FLUPZRZJKs8Vk\n\t+RclOakx4dCWNRgK+DROH70T+OnNP2kZHBsmI97UtiIl75hpXxfn+Jcicm18OrssiPP1Q379Aay\n\t61apN0HbeA=","X-Received":"by 2002:a05:7300:371f:b0:2c5:b23e:48a6 with SMTP id\n 5a478bee46e88-2cbfc268d12mr658562eec.23.1775177510944;\n Thu, 02 Apr 2026 17:51:50 -0700 (PDT)","From":"Douglas Anderson ","To":"Greg Kroah-Hartman ,\n\t\"Rafael J . Wysocki\" ,\n\tDanilo Krummrich ,\n\tAlan Stern ","Cc":"Robin Murphy ,\n\tLeon Romanovsky ,\n\tPaul Burton ,\n\tSaravana Kannan ,\n\tAlexander Lobakin ,\n\tEric Dumazet ,\n\tToshi Kani ,\n\tChristoph Hellwig ,\n\tAlexey Kardashevskiy ,\n\tJohan Hovold ,\n\tDouglas Anderson ,\n\tchleroy@kernel.org,\n\tdriver-core@lists.linux.dev,\n\tgbatra@linux.ibm.com,\n\tiommu@lists.linux.dev,\n\tlinux-kernel@vger.kernel.org,\n\tlinuxppc-dev@lists.ozlabs.org,\n\tm.szyprowski@samsung.com,\n\tmaddy@linux.ibm.com,\n\tmpe@ellerman.id.au,\n\tnpiggin@gmail.com","Subject":"[PATCH v3 5/9] driver core: Replace dev->dma_ops_bypass with\n DEV_FLAG_DMA_OPS_BYPASS","Date":"Thu, 2 Apr 2026 17:49:51 -0700","Message-ID":"\n <20260402174925.v3.5.If62b84471ef2c85e7ad250f0468867d6dba965ab@changeid>","X-Mailer":"git-send-email 2.53.0.1213.gd9a14994de-goog","In-Reply-To":"<20260403005005.30424-1-dianders@chromium.org>","References":"<20260403005005.30424-1-dianders@chromium.org>","X-Mailing-List":"linuxppc-dev@lists.ozlabs.org","List-Id":"","List-Help":"","List-Owner":"","List-Post":"","List-Archive":",\n ","List-Subscribe":",\n ,\n ","List-Unsubscribe":"","Precedence":"list","MIME-Version":"1.0","Content-Transfer-Encoding":"8bit","X-Spam-Status":"No, score=-0.2 required=3.0 tests=DKIMWL_WL_HIGH,DKIM_SIGNED,\n\tDKIM_VALID,DKIM_VALID_AU,DKIM_VALID_EF,RCVD_IN_DNSWL_NONE,\n\tSPF_HELO_NONE,SPF_PASS autolearn=disabled version=4.0.1 OzLabs 8","X-Spam-Checker-Version":"SpamAssassin 4.0.1 (2024-03-25) on lists.ozlabs.org"},"content":"In C, bitfields are not necessarily safe to modify from multiple\nthreads without locking. Switch \"dma_ops_bypass\" over to the \"flags\"\nfield so modifications are safe.\n\nCc: Christoph Hellwig \nCc: Alexey Kardashevskiy \nSigned-off-by: Douglas Anderson \n---\nNot fixing any known bugs; problem is theoretical and found by code\ninspection. Change is done somewhat manually and only lightly tested\n(mostly compile-time tested).\n\nNOTE: even though previously we only took up a bit if\nCONFIG_DMA_OPS_BYPASS, in this change I reserve the bit\nunconditionally. While we could get the \"dynamic\" behavior by\nchanging the flags definition to be an \"enum\", it doesn't seem worth\nit at this point. This also allows us to move one \"#ifdef\" to an \"if\",\ngetting better compile-time testing of both sides of the \"if\".\n\nChanges in v3:\n- New\n\n arch/powerpc/kernel/dma-iommu.c | 8 ++++----\n include/linux/device.h | 14 ++++++--------\n kernel/dma/mapping.c | 8 +++-----\n 3 files changed, 13 insertions(+), 17 deletions(-)","diff":"diff --git a/arch/powerpc/kernel/dma-iommu.c b/arch/powerpc/kernel/dma-iommu.c\nindex 73e10bd4d56d..6486417c8acb 100644\n--- a/arch/powerpc/kernel/dma-iommu.c\n+++ b/arch/powerpc/kernel/dma-iommu.c\n@@ -67,7 +67,7 @@ bool arch_dma_unmap_sg_direct(struct device *dev, struct scatterlist *sg,\n }\n bool arch_dma_alloc_direct(struct device *dev)\n {\n-\tif (dev->dma_ops_bypass)\n+\tif (test_bit(DEV_FLAG_DMA_OPS_BYPASS, &dev->flags))\n \t\treturn true;\n \n \treturn false;\n@@ -75,7 +75,7 @@ bool arch_dma_alloc_direct(struct device *dev)\n \n bool arch_dma_free_direct(struct device *dev, dma_addr_t dma_handle)\n {\n-\tif (!dev->dma_ops_bypass)\n+\tif (!test_bit(DEV_FLAG_DMA_OPS_BYPASS, &dev->flags))\n \t\treturn false;\n \n \treturn is_direct_handle(dev, dma_handle);\n@@ -164,7 +164,7 @@ int dma_iommu_dma_supported(struct device *dev, u64 mask)\n \t\t * fixed ops will be used for RAM. This is limited by\n \t\t * bus_dma_limit which is set when RAM is pre-mapped.\n \t\t */\n-\t\tdev->dma_ops_bypass = true;\n+\t\tset_bit(DEV_FLAG_DMA_OPS_BYPASS, &dev->flags);\n \t\tdev_info(dev, \"iommu: 64-bit OK but direct DMA is limited by %llx\\n\",\n \t\t\t dev->bus_dma_limit);\n \t\treturn 1;\n@@ -185,7 +185,7 @@ int dma_iommu_dma_supported(struct device *dev, u64 mask)\n \t}\n \n \tdev_dbg(dev, \"iommu: not 64-bit, using default ops\\n\");\n-\tdev->dma_ops_bypass = false;\n+\tclear_bit(DEV_FLAG_DMA_OPS_BYPASS, &dev->flags);\n \treturn 1;\n }\n \ndiff --git a/include/linux/device.h b/include/linux/device.h\nindex e900748d3038..f5845bd7c3e6 100644\n--- a/include/linux/device.h\n+++ b/include/linux/device.h\n@@ -472,12 +472,18 @@ struct device_physical_location {\n *\t\tdoesn't rely on dma_ops structure.\n * @DEV_FLAG_DMA_SKIP_SYNC: DMA sync operations can be skipped for coherent\n *\t\tbuffers.\n+ * @DEV_FLAG_DMA_OPS_BYPASS: If set then the dma_ops are bypassed for the\n+ *\t\tstreaming DMA operations (->map_* / ->unmap_* / ->sync_*), and\n+ *\t\toptional (if the coherent mask is large enough) also for dma\n+ *\t\tallocations. This flag is managed by the dma ops instance from\n+ *\t\t->dma_supported.\n */\n enum struct_device_flags {\n \tDEV_FLAG_READY_TO_PROBE,\n \tDEV_FLAG_CAN_MATCH,\n \tDEV_FLAG_DMA_IOMMU,\n \tDEV_FLAG_DMA_SKIP_SYNC,\n+\tDEV_FLAG_DMA_OPS_BYPASS,\n };\n \n /**\n@@ -564,11 +570,6 @@ enum struct_device_flags {\n *\t\t sync_state() callback.\n * @dma_coherent: this particular device is dma coherent, even if the\n *\t\tarchitecture supports non-coherent devices.\n- * @dma_ops_bypass: If set to %true then the dma_ops are bypassed for the\n- *\t\tstreaming DMA operations (->map_* / ->unmap_* / ->sync_*),\n- *\t\tand optionall (if the coherent mask is large enough) also\n- *\t\tfor dma allocations. This flag is managed by the dma ops\n- *\t\tinstance from ->dma_supported.\n * @flags:\tDEV_FLAG_XXX flags. Use atomic bitfield operations to modify.\n *\n * At the lowest level, every device in a Linux system is represented by an\n@@ -682,9 +683,6 @@ struct device {\n defined(CONFIG_ARCH_HAS_SYNC_DMA_FOR_CPU_ALL)\n \tbool\t\t\tdma_coherent:1;\n #endif\n-#ifdef CONFIG_DMA_OPS_BYPASS\n-\tbool\t\t\tdma_ops_bypass : 1;\n-#endif\n \n \tunsigned long\t\tflags;\n };\ndiff --git a/kernel/dma/mapping.c b/kernel/dma/mapping.c\nindex f50b648ed460..44d4c319ffad 100644\n--- a/kernel/dma/mapping.c\n+++ b/kernel/dma/mapping.c\n@@ -126,11 +126,9 @@ static bool dma_go_direct(struct device *dev, dma_addr_t mask,\n \tif (likely(!ops))\n \t\treturn true;\n \n-#ifdef CONFIG_DMA_OPS_BYPASS\n-\tif (dev->dma_ops_bypass)\n+\tif (IS_ENABLED(CONFIG_DMA_OPS_BYPASS) && test_bit(DEV_FLAG_DMA_OPS_BYPASS, &dev->flags))\n \t\treturn min_not_zero(mask, dev->bus_dma_limit) >=\n \t\t\t dma_direct_get_required_mask(dev);\n-#endif\n \treturn false;\n }\n \n@@ -895,8 +893,8 @@ bool dma_pci_p2pdma_supported(struct device *dev)\n \tconst struct dma_map_ops *ops = get_dma_ops(dev);\n \n \t/*\n-\t * Note: dma_ops_bypass is not checked here because P2PDMA should\n-\t * not be used with dma mapping ops that do not have support even\n+\t * Note: DEV_FLAG_DMA_OPS_BYPASS is not checked here because P2PDMA\n+\t * should not be used with dma mapping ops that do not have support even\n \t * if the specific device is bypassing them.\n \t */\n \n","prefixes":["v3","5/9"]}