Move the code from the inline functions into exported functions.

Signed-off-by: Christian König christian.koenig@amd.com Acked-by: Daniel Vetter daniel.vetter@ffwll.ch --- drivers/dma-buf/Makefile | 2 +- drivers/dma-buf/dma-fence-unwrap.c | 59 ++++++++++++++++++++++++++++++ include/linux/dma-fence-unwrap.h | 52 ++------------------------ 3 files changed, 64 insertions(+), 49 deletions(-) create mode 100644 drivers/dma-buf/dma-fence-unwrap.c

diff --git a/drivers/dma-buf/Makefile b/drivers/dma-buf/Makefile index 4c9eb53ba3f8..70ec901edf2c 100644 --- a/drivers/dma-buf/Makefile +++ b/drivers/dma-buf/Makefile @@ -1,6 +1,6 @@ # SPDX-License-Identifier: GPL-2.0-only obj-y := dma-buf.o dma-fence.o dma-fence-array.o dma-fence-chain.o \ - dma-resv.o + dma-fence-unwrap.o dma-resv.o obj-$(CONFIG_DMABUF_HEAPS) += dma-heap.o obj-$(CONFIG_DMABUF_HEAPS) += heaps/ obj-$(CONFIG_SYNC_FILE) += sync_file.o diff --git a/drivers/dma-buf/dma-fence-unwrap.c b/drivers/dma-buf/dma-fence-unwrap.c new file mode 100644 index 000000000000..711be125428c --- /dev/null +++ b/drivers/dma-buf/dma-fence-unwrap.c @@ -0,0 +1,59 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * dma-fence-util: misc functions for dma_fence objects + * + * Copyright (C) 2022 Advanced Micro Devices, Inc. + * Authors: + * Christian König christian.koenig@amd.com + */ + +#include <linux/dma-fence.h> +#include <linux/dma-fence-array.h> +#include <linux/dma-fence-chain.h> +#include <linux/dma-fence-unwrap.h> + +/* Internal helper to start new array iteration, don't use directly */ +static struct dma_fence * +__dma_fence_unwrap_array(struct dma_fence_unwrap *cursor) +{ + cursor->array = dma_fence_chain_contained(cursor->chain); + cursor->index = 0; + return dma_fence_array_first(cursor->array); +} + +/** + * dma_fence_unwrap_first - return the first fence from fence containers + * @head: the entrypoint into the containers + * @cursor: current position inside the containers + * + * Unwraps potential dma_fence_chain/dma_fence_array containers and return the + * first fence. + */ +struct dma_fence *dma_fence_unwrap_first(struct dma_fence *head, + struct dma_fence_unwrap *cursor) +{ + cursor->chain = dma_fence_get(head); + return __dma_fence_unwrap_array(cursor); +} +EXPORT_SYMBOL_GPL(dma_fence_unwrap_first); + +/** + * dma_fence_unwrap_next - return the next fence from a fence containers + * @cursor: current position inside the containers + * + * Continue unwrapping the dma_fence_chain/dma_fence_array containers and return + * the next fence from them. + */ +struct dma_fence *dma_fence_unwrap_next(struct dma_fence_unwrap *cursor) +{ + struct dma_fence *tmp; + + ++cursor->index; + tmp = dma_fence_array_next(cursor->array, cursor->index); + if (tmp) + return tmp; + + cursor->chain = dma_fence_chain_walk(cursor->chain); + return __dma_fence_unwrap_array(cursor); +} +EXPORT_SYMBOL_GPL(dma_fence_unwrap_next); diff --git a/include/linux/dma-fence-unwrap.h b/include/linux/dma-fence-unwrap.h index 77e335a1bcac..e7c219da4ed7 100644 --- a/include/linux/dma-fence-unwrap.h +++ b/include/linux/dma-fence-unwrap.h @@ -1,7 +1,5 @@ /* SPDX-License-Identifier: GPL-2.0-only */ /* - * fence-chain: chain fences together in a timeline - * * Copyright (C) 2022 Advanced Micro Devices, Inc. * Authors: * Christian König christian.koenig@amd.com @@ -10,8 +8,7 @@ #ifndef __LINUX_DMA_FENCE_UNWRAP_H #define __LINUX_DMA_FENCE_UNWRAP_H

-#include <linux/dma-fence-chain.h> -#include <linux/dma-fence-array.h> +struct dma_fence;

/** * struct dma_fence_unwrap - cursor into the container structure @@ -33,50 +30,9 @@ struct dma_fence_unwrap { unsigned int index; };

-/* Internal helper to start new array iteration, don't use directly */ -static inline struct dma_fence * -__dma_fence_unwrap_array(struct dma_fence_unwrap * cursor) -{ - cursor->array = dma_fence_chain_contained(cursor->chain); - cursor->index = 0; - return dma_fence_array_first(cursor->array); -} - -/** - * dma_fence_unwrap_first - return the first fence from fence containers - * @head: the entrypoint into the containers - * @cursor: current position inside the containers - * - * Unwraps potential dma_fence_chain/dma_fence_array containers and return the - * first fence. - */ -static inline struct dma_fence * -dma_fence_unwrap_first(struct dma_fence *head, struct dma_fence_unwrap *cursor) -{ - cursor->chain = dma_fence_get(head); - return __dma_fence_unwrap_array(cursor); -} - -/** - * dma_fence_unwrap_next - return the next fence from a fence containers - * @cursor: current position inside the containers - * - * Continue unwrapping the dma_fence_chain/dma_fence_array containers and return - * the next fence from them. - */ -static inline struct dma_fence * -dma_fence_unwrap_next(struct dma_fence_unwrap *cursor) -{ - struct dma_fence *tmp; - - ++cursor->index; - tmp = dma_fence_array_next(cursor->array, cursor->index); - if (tmp) - return tmp; - - cursor->chain = dma_fence_chain_walk(cursor->chain); - return __dma_fence_unwrap_array(cursor); -} +struct dma_fence *dma_fence_unwrap_first(struct dma_fence *head, + struct dma_fence_unwrap *cursor); +struct dma_fence *dma_fence_unwrap_next(struct dma_fence_unwrap *cursor);

/** * dma_fence_unwrap_for_each - iterate over all fences in containers

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 christian.koenig@amd.com Reviewed-by: Daniel Vetter daniel.vetter@ffwll.ch --- 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 <linux/dma-fence-array.h> #include <linux/dma-fence-chain.h> #include <linux/dma-fence-unwrap.h> +#include <linux/slab.h>

/* 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

The unwrap merge function is now intended for this use case.

Signed-off-by: Christian König christian.koenig@amd.com Reviewed-by: Daniel Vetter daniel.vetter@ffwll.ch --- drivers/gpu/drm/drm_syncobj.c | 57 +++++------------------------------ 1 file changed, 7 insertions(+), 50 deletions(-)

diff --git a/drivers/gpu/drm/drm_syncobj.c b/drivers/gpu/drm/drm_syncobj.c index 7e48dcd1bee4..bbad9e4696e7 100644 --- a/drivers/gpu/drm/drm_syncobj.c +++ b/drivers/gpu/drm/drm_syncobj.c @@ -184,6 +184,7 @@ */

#include <linux/anon_inodes.h> +#include <linux/dma-fence-unwrap.h> #include <linux/file.h> #include <linux/fs.h> #include <linux/sched/signal.h> @@ -853,57 +854,12 @@ drm_syncobj_fd_to_handle_ioctl(struct drm_device *dev, void *data, &args->handle); }

- -/* - * Try to flatten a dma_fence_chain into a dma_fence_array so that it can be - * added as timeline fence to a chain again. - */ -static int drm_syncobj_flatten_chain(struct dma_fence **f) -{ - struct dma_fence_chain *chain = to_dma_fence_chain(*f); - struct dma_fence *tmp, **fences; - struct dma_fence_array *array; - unsigned int count; - - if (!chain) - return 0; - - count = 0; - dma_fence_chain_for_each(tmp, &chain->base) - ++count; - - fences = kmalloc_array(count, sizeof(*fences), GFP_KERNEL); - if (!fences) - return -ENOMEM; - - count = 0; - dma_fence_chain_for_each(tmp, &chain->base) - fences[count++] = dma_fence_get(tmp); - - array = dma_fence_array_create(count, fences, - dma_fence_context_alloc(1), - 1, false); - if (!array) - goto free_fences; - - dma_fence_put(*f); - *f = &array->base; - return 0; - -free_fences: - while (count--) - dma_fence_put(fences[count]); - - kfree(fences); - return -ENOMEM; -} - static int drm_syncobj_transfer_to_timeline(struct drm_file *file_private, struct drm_syncobj_transfer *args) { struct drm_syncobj *timeline_syncobj = NULL; + struct dma_fence *fence, *tmp; struct dma_fence_chain *chain; - struct dma_fence *fence; int ret;

timeline_syncobj = drm_syncobj_find(file_private, args->dst_handle); @@ -912,13 +868,14 @@ static int drm_syncobj_transfer_to_timeline(struct drm_file *file_private, } ret = drm_syncobj_find_fence(file_private, args->src_handle, args->src_point, args->flags, - &fence); + &tmp); if (ret) goto err_put_timeline;

- ret = drm_syncobj_flatten_chain(&fence); - if (ret) - goto err_free_fence; + fence = dma_fence_unwrap_merge(tmp); + dma_fence_put(tmp); + if (!fence) + goto err_put_timeline;

chain = dma_fence_chain_alloc(); if (!chain) {