From patchwork Fri May 6 14:10:08 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: =?utf-8?q?Christian_K=C3=B6nig?= X-Patchwork-Id: 12841238 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 DBFDCC433EF for ; Fri, 6 May 2022 14:10:20 +0000 (UTC) Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 6422910EE3F; Fri, 6 May 2022 14:10:18 +0000 (UTC) Received: from mail-ed1-x52d.google.com (mail-ed1-x52d.google.com [IPv6:2a00:1450:4864:20::52d]) by gabe.freedesktop.org (Postfix) with ESMTPS id 661F110ED8B for ; Fri, 6 May 2022 14:10:16 +0000 (UTC) Received: by mail-ed1-x52d.google.com with SMTP id p18so8854420edr.7 for ; Fri, 06 May 2022 07:10:16 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=vefTnLmZ1hRjfFT+wM4LObjotAMbQOwWtxRJQRE0yXw=; b=j8KVGH2GHcMJ1ywj+cjz7pamBQLXExx8mmlGN6PbeSlm/kg++H2qKPy19AUZrDjN0Z o8FHJEp9fS1xdrFOQhSl9cXB2IJU4D7vQlMOewun/H7B6oHQ11No5Scm2xOpkcxDdT7H 1TECQagWPIl96unP5+2bTbHhUnBnRS8y3kLYliVxg05CyWXc1xqSMVIBefKaUQMGvI8B 4YYizgE1ioczKXABvNTntR0tKAty5FnafWl08Aw1pDKl3Wc6q3ZJjsBZHhA46HU8sflS TMcKOlIP0TG4W2/GqjtjLxHsjwvXYlmG4ZhVmO/TKJjur5dp/UnDEv2B/+zc40e/yr4p h7gw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=vefTnLmZ1hRjfFT+wM4LObjotAMbQOwWtxRJQRE0yXw=; b=GK84Wl8OL4doe45V3lbqzHLTXPEHQ4lyfn4a9JPB6a1erXa6vyGsXL+rXOLLHzGNsJ NKftkeWW95Uyl8+ATGtztj1lVifVoL19UoJmjob6TQCVTfbzMKQlku8hETwT6EvWl4QC re1gzIwtbf1xx8Xgni2uutJQLvwS0e9YVt2cTXXpL1TeM2s5kSSuOyc3jO5dOKN//Rz3 5Dm06X1kqnuy9fes2vkBVxYriAQhMdmZ85lZLmA8h5aynKYn1a0S3H3BdEow9xw3sR/T Gddd3FKU+gBMlYTNvYiid8U5SbrpbTz2HKj47DQ97g8marZStB3LaGNtJCZFgA9Gp0k/ Xjdw== X-Gm-Message-State: AOAM5319n2V40J93DAM0/lxyhyFkS8cCP9uMN43XL3wlw+q4bSaRkYGj td6QylgAc0+HTBGFtUV7d0U= X-Google-Smtp-Source: ABdhPJwxAnRnqAaKfldxjBs/gMSK370De63PnkL12J/y9NEBXjJENsVHF+C8jsEr0H/sNCoxeRjgGw== X-Received: by 2002:a50:ed13:0:b0:426:4939:45a9 with SMTP id j19-20020a50ed13000000b00426493945a9mr3701112eds.303.1651846214835; Fri, 06 May 2022 07:10:14 -0700 (PDT) Received: from able.fritz.box (p57b0b3fd.dip0.t-ipconnect.de. [87.176.179.253]) by smtp.gmail.com with ESMTPSA id jx3-20020a170907760300b006f3ef214e22sm1915415ejc.136.2022.05.06.07.10.13 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 06 May 2022 07:10:14 -0700 (PDT) From: " =?utf-8?q?Christian_K=C3=B6nig?= " X-Google-Original-From: =?utf-8?q?Christian_K=C3=B6nig?= To: daniel@ffwll.ch, linaro-mm-sig@lists.linaro.org, dri-devel@lists.freedesktop.org, linux-media@vger.kernel.org Subject: [PATCH 4/5] dma-buf: generalize dma_fence unwrap & merging v2 Date: Fri, 6 May 2022 16:10:08 +0200 Message-Id: <20220506141009.18047-4-christian.koenig@amd.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20220506141009.18047-1-christian.koenig@amd.com> References: <20220506141009.18047-1-christian.koenig@amd.com> MIME-Version: 1.0 X-BeenThere: dri-devel@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Direct Rendering Infrastructure - Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Daniel Vetter , =?utf-8?q?Christian_K=C3=B6nig?= Errors-To: dri-devel-bounces@lists.freedesktop.org Sender: "dri-devel" Introduce a dma_fence_unwrap_merge() macro which allows to unwrap fences which potentially can be containers as well and then merge them back together into a flat dma_fence_array. v2: rename the function, add some more comments about how the wrapper is used, move filtering of signaled fences into the unwrap iterator, add complex selftest which covers more cases. Signed-off-by: Christian König Reviewed-by: Daniel Vetter Reported-by: kernel test robot --- drivers/dma-buf/dma-fence-unwrap.c | 99 +++++++++++++++++++++ drivers/dma-buf/st-dma-fence-unwrap.c | 109 +++++++++++++++++++++++ drivers/dma-buf/sync_file.c | 119 ++------------------------ include/linux/dma-fence-unwrap.h | 24 ++++++ 4 files changed, 238 insertions(+), 113 deletions(-) diff --git a/drivers/dma-buf/dma-fence-unwrap.c b/drivers/dma-buf/dma-fence-unwrap.c index 711be125428c..0e51547bbabd 100644 --- a/drivers/dma-buf/dma-fence-unwrap.c +++ b/drivers/dma-buf/dma-fence-unwrap.c @@ -11,6 +11,7 @@ #include #include #include +#include /* Internal helper to start new array iteration, don't use directly */ static struct dma_fence * @@ -57,3 +58,101 @@ struct dma_fence *dma_fence_unwrap_next(struct dma_fence_unwrap *cursor) return __dma_fence_unwrap_array(cursor); } EXPORT_SYMBOL_GPL(dma_fence_unwrap_next); + +/* Implementation for the dma_fence_merge() marco, don't use directly */ +struct dma_fence *__dma_fence_unwrap_merge(unsigned int num_fences, + struct dma_fence **fences, + struct dma_fence_unwrap *iter) +{ + struct dma_fence_array *result; + struct dma_fence *tmp, **array; + unsigned int i; + size_t count; + + count = 0; + for (i = 0; i < num_fences; ++i) { + dma_fence_unwrap_for_each(tmp, &iter[i], fences[i]) + ++count; + } + + if (count == 0) + return dma_fence_get_stub(); + + array = kmalloc_array(count, sizeof(*array), GFP_KERNEL); + if (!array) + return NULL; + + /* + * This trashes the input fence array and uses it as position for the + * following merge loop. This works because the dma_fence_merge() + * wrapper macro is creating this temporary array on the stack together + * with the iterators. + */ + for (i = 0; i < num_fences; ++i) + fences[i] = dma_fence_unwrap_first(fences[i], &iter[i]); + + count = 0; + do { + unsigned int sel; + +restart: + tmp = NULL; + for (i = 0; i < num_fences; ++i) { + struct dma_fence *next = fences[i]; + + if (!next) + continue; + + /* + * We can't guarantee that inpute fences are ordered by + * context, but it is still quite likely when this + * function is used multiple times. So attempt to order + * the fences by context as we pass over them and merge + * fences with the same context. + */ + if (!tmp || tmp->context > next->context) { + tmp = next; + sel = i; + + } else if (tmp->context < next->context) { + continue; + + } else if (dma_fence_is_later(tmp, next)) { + fences[i] = dma_fence_unwrap_next(&iter[i]); + goto restart; + } else { + fences[sel] = dma_fence_unwrap_next(&iter[sel]); + goto restart; + } + } + + if (tmp) { + array[count++] = dma_fence_get(tmp); + fences[sel] = dma_fence_unwrap_next(&iter[sel]); + } + } while (tmp); + + if (count == 0) { + tmp = dma_fence_get_stub(); + goto return_tmp; + } + + if (count == 1) { + tmp = array[0]; + goto return_tmp; + } + + result = dma_fence_array_create(count, array, + dma_fence_context_alloc(1), + 1, false); + if (!result) { + tmp = NULL; + goto return_tmp; + } + return &result->base; + +return_tmp: + kfree(array); + return tmp; +} +EXPORT_SYMBOL_GPL(__dma_fence_unwrap_merge); diff --git a/drivers/dma-buf/st-dma-fence-unwrap.c b/drivers/dma-buf/st-dma-fence-unwrap.c index e20c5a7dcfe4..4105d5ea8dde 100644 --- a/drivers/dma-buf/st-dma-fence-unwrap.c +++ b/drivers/dma-buf/st-dma-fence-unwrap.c @@ -238,6 +238,113 @@ static int unwrap_chain_array(void *arg) return err; } +static int unwrap_merge(void *arg) +{ + struct dma_fence *fence, *f1, *f2, *f3; + struct dma_fence_unwrap iter; + int err = 0; + + f1 = mock_fence(); + if (!f1) + return -ENOMEM; + + f2 = mock_fence(); + if (!f2) { + err = -ENOMEM; + goto error_put_f1; + } + + f3 = dma_fence_unwrap_merge(f1, f2); + if (!f3) { + err = -ENOMEM; + goto error_put_f2; + } + + dma_fence_unwrap_for_each(fence, &iter, f3) { + if (fence == f1) { + dma_fence_put(f1); + f1 = NULL; + } else if (fence == f2) { + dma_fence_put(f2); + f2 = NULL; + } else { + pr_err("Unexpected fence!\n"); + err = -EINVAL; + } + } + + if (f1 || f2) { + pr_err("Not all fences seen!\n"); + err = -EINVAL; + } + + dma_fence_put(f3); +error_put_f2: + dma_fence_put(f2); +error_put_f1: + dma_fence_put(f1); + return err; +} + +static int unwrap_merge_complex(void *arg) +{ + struct dma_fence *fence, *f1, *f2, *f3, *f4, *f5; + struct dma_fence_unwrap iter; + int err = -ENOMEM; + + f1 = mock_fence(); + if (!f1) + return -ENOMEM; + + f2 = mock_fence(); + if (!f2) + goto error_put_f1; + + f3 = dma_fence_unwrap_merge(f1, f2); + if (!f3) + goto error_put_f2; + + /* The resulting array has the fences in reverse */ + f4 = dma_fence_unwrap_merge(f2, f1); + if (!f4) + goto error_put_f3; + + /* Signaled fences should be filtered, the two arrays merged. */ + f5 = dma_fence_unwrap_merge(f3, f4, dma_fence_get_stub()); + if (!f5) + goto error_put_f4; + + err = 0; + dma_fence_unwrap_for_each(fence, &iter, f5) { + if (fence == f1) { + dma_fence_put(f1); + f1 = NULL; + } else if (fence == f2) { + dma_fence_put(f2); + f2 = NULL; + } else { + pr_err("Unexpected fence!\n"); + err = -EINVAL; + } + } + + if (f1 || f2) { + pr_err("Not all fences seen!\n"); + err = -EINVAL; + } + + dma_fence_put(f5); +error_put_f4: + dma_fence_put(f4); +error_put_f3: + dma_fence_put(f3); +error_put_f2: + dma_fence_put(f2); +error_put_f1: + dma_fence_put(f1); + return err; +} + int dma_fence_unwrap(void) { static const struct subtest tests[] = { @@ -245,6 +352,8 @@ int dma_fence_unwrap(void) SUBTEST(unwrap_array), SUBTEST(unwrap_chain), SUBTEST(unwrap_chain_array), + SUBTEST(unwrap_merge), + SUBTEST(unwrap_merge_complex), }; return subtests(tests, NULL); diff --git a/drivers/dma-buf/sync_file.c b/drivers/dma-buf/sync_file.c index 0fe564539166..3ebec19a8e02 100644 --- a/drivers/dma-buf/sync_file.c +++ b/drivers/dma-buf/sync_file.c @@ -146,50 +146,6 @@ char *sync_file_get_name(struct sync_file *sync_file, char *buf, int len) return buf; } -static int sync_file_set_fence(struct sync_file *sync_file, - struct dma_fence **fences, int num_fences) -{ - struct dma_fence_array *array; - - /* - * The reference for the fences in the new sync_file and held - * in add_fence() during the merge procedure, so for num_fences == 1 - * we already own a new reference to the fence. For num_fence > 1 - * we own the reference of the dma_fence_array creation. - */ - - if (num_fences == 0) { - sync_file->fence = dma_fence_get_stub(); - kfree(fences); - - } else if (num_fences == 1) { - sync_file->fence = fences[0]; - kfree(fences); - - } else { - array = dma_fence_array_create(num_fences, fences, - dma_fence_context_alloc(1), - 1, false); - if (!array) - return -ENOMEM; - - sync_file->fence = &array->base; - } - - return 0; -} - -static void add_fence(struct dma_fence **fences, - int *i, struct dma_fence *fence) -{ - fences[*i] = fence; - - if (!dma_fence_is_signaled(fence)) { - dma_fence_get(fence); - (*i)++; - } -} - /** * sync_file_merge() - merge two sync_files * @name: name of new fence @@ -203,84 +159,21 @@ static void add_fence(struct dma_fence **fences, static struct sync_file *sync_file_merge(const char *name, struct sync_file *a, struct sync_file *b) { - struct dma_fence *a_fence, *b_fence, **fences; - struct dma_fence_unwrap a_iter, b_iter; - unsigned int index, num_fences; struct sync_file *sync_file; + struct dma_fence *fence; sync_file = sync_file_alloc(); if (!sync_file) return NULL; - num_fences = 0; - dma_fence_unwrap_for_each(a_fence, &a_iter, a->fence) - ++num_fences; - dma_fence_unwrap_for_each(b_fence, &b_iter, b->fence) - ++num_fences; - - if (num_fences > INT_MAX) - goto err_free_sync_file; - - fences = kcalloc(num_fences, sizeof(*fences), GFP_KERNEL); - if (!fences) - goto err_free_sync_file; - - /* - * We can't guarantee that fences in both a and b are ordered, but it is - * still quite likely. - * - * So attempt to order the fences as we pass over them and merge fences - * with the same context. - */ - - index = 0; - for (a_fence = dma_fence_unwrap_first(a->fence, &a_iter), - b_fence = dma_fence_unwrap_first(b->fence, &b_iter); - a_fence || b_fence; ) { - - if (!b_fence) { - add_fence(fences, &index, a_fence); - a_fence = dma_fence_unwrap_next(&a_iter); - - } else if (!a_fence) { - add_fence(fences, &index, b_fence); - b_fence = dma_fence_unwrap_next(&b_iter); - - } else if (a_fence->context < b_fence->context) { - add_fence(fences, &index, a_fence); - a_fence = dma_fence_unwrap_next(&a_iter); - - } else if (b_fence->context < a_fence->context) { - add_fence(fences, &index, b_fence); - b_fence = dma_fence_unwrap_next(&b_iter); - - } else if (__dma_fence_is_later(a_fence->seqno, b_fence->seqno, - a_fence->ops)) { - add_fence(fences, &index, a_fence); - a_fence = dma_fence_unwrap_next(&a_iter); - b_fence = dma_fence_unwrap_next(&b_iter); - - } else { - add_fence(fences, &index, b_fence); - a_fence = dma_fence_unwrap_next(&a_iter); - b_fence = dma_fence_unwrap_next(&b_iter); - } + fence = dma_fence_unwrap_merge(a->fence, b->fence); + if (!fence) { + fput(sync_file->file); + return NULL; } - - if (sync_file_set_fence(sync_file, fences, index) < 0) - goto err_put_fences; - + sync_file->fence = fence; strlcpy(sync_file->user_name, name, sizeof(sync_file->user_name)); return sync_file; - -err_put_fences: - while (index) - dma_fence_put(fences[--index]); - kfree(fences); - -err_free_sync_file: - fput(sync_file->file); - return NULL; } static int sync_file_release(struct inode *inode, struct file *file) diff --git a/include/linux/dma-fence-unwrap.h b/include/linux/dma-fence-unwrap.h index a4d342fef8e0..390de1ee9d35 100644 --- a/include/linux/dma-fence-unwrap.h +++ b/include/linux/dma-fence-unwrap.h @@ -52,4 +52,28 @@ struct dma_fence *dma_fence_unwrap_next(struct dma_fence_unwrap *cursor); fence = dma_fence_unwrap_next(cursor)) \ if (!dma_fence_is_signaled(fence)) +struct dma_fence *__dma_fence_unwrap_merge(unsigned int num_fences, + struct dma_fence **fences, + struct dma_fence_unwrap *cursors); + +/** + * dma_fence_unwrap_merge - unwrap and merge fences + * + * All fences given as parameters are unwrapped and merged back together as flat + * dma_fence_array. Useful if multiple containers need to be merged together. + * + * Implemented as a macro to allocate the necessary arrays on the stack and + * account the stack frame size to the caller. + * + * Returns NULL on memory allocation failure, a dma_fence object representing + * all the given fences otherwise. + */ +#define dma_fence_unwrap_merge(...) \ + ({ \ + struct dma_fence *__f[] = { __VA_ARGS__ }; \ + struct dma_fence_unwrap __c[ARRAY_SIZE(__f)]; \ + \ + __dma_fence_unwrap_merge(ARRAY_SIZE(__f), __f, __c); \ + }) + #endif