From d637c17832bd020c27b3c29c0bd27275d1af9132 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Tue, 6 Dec 2016 12:40:51 +0000 Subject: [PATCH] drm/i915: Use memcpy_from_wc for GPU error capture On all platforms we now always read the contents of buffers via the GTT, i.e. using WC cpu access. Reads are slow, but they can be accelerated with an internal read buffer using sse4.1 (movntqda). This is our i915_memcpy_from_wc() routine which also checks for sse4.1 support and so we can fallback to using a regular slow memcpy if we need to. When compressing the pages, the reads are currently done inside zlib's fill_window() routine and so we must copy the page into a temporary which is then already inside the CPU cache and fast for zlib's compression. When not compressing the pages, we don't need a temporary and can just use the accelerated read from WC into the destination. v2: Use zstream locals to reduce diff and allocate the additional temporary storage only if sse4.1 is supported. v3: Use length=0 for the sse4.1 support check Signed-off-by: Chris Wilson Cc: Tvrtko Ursulin Link: http://patchwork.freedesktop.org/patch/msgid/20161206124051.17040-1-chris@chris-wilson.co.uk Reviewed-by: Tvrtko Ursulin --- drivers/gpu/drm/i915/i915_gpu_error.c | 50 ++++++++++++++++++++------- 1 file changed, 37 insertions(+), 13 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c index a14f7badc337..307999b852fd 100644 --- a/drivers/gpu/drm/i915/i915_gpu_error.c +++ b/drivers/gpu/drm/i915/i915_gpu_error.c @@ -176,9 +176,14 @@ static void i915_error_puts(struct drm_i915_error_state_buf *e, #ifdef CONFIG_DRM_I915_COMPRESS_ERROR -static bool compress_init(struct z_stream_s *zstream) +struct compress { + struct z_stream_s zstream; + void *tmp; +}; + +static bool compress_init(struct compress *c) { - memset(zstream, 0, sizeof(*zstream)); + struct z_stream_s *zstream = memset(&c->zstream, 0, sizeof(c->zstream)); zstream->workspace = kmalloc(zlib_deflate_workspacesize(MAX_WBITS, MAX_MEM_LEVEL), @@ -191,14 +196,22 @@ static bool compress_init(struct z_stream_s *zstream) return false; } + c->tmp = NULL; + if (i915_memcpy_from_wc(NULL, 0, 0)) + c->tmp = (void *)__get_free_page(GFP_ATOMIC | __GFP_NOWARN); + return true; } -static int compress_page(struct z_stream_s *zstream, +static int compress_page(struct compress *c, void *src, struct drm_i915_error_object *dst) { + struct z_stream_s *zstream = &c->zstream; + zstream->next_in = src; + if (c->tmp && i915_memcpy_from_wc(c->tmp, src, PAGE_SIZE)) + zstream->next_in = c->tmp; zstream->avail_in = PAGE_SIZE; do { @@ -226,9 +239,11 @@ static int compress_page(struct z_stream_s *zstream, return 0; } -static void compress_fini(struct z_stream_s *zstream, +static void compress_fini(struct compress *c, struct drm_i915_error_object *dst) { + struct z_stream_s *zstream = &c->zstream; + if (dst) { zlib_deflate(zstream, Z_FINISH); dst->unused = zstream->avail_out; @@ -236,6 +251,9 @@ static void compress_fini(struct z_stream_s *zstream, zlib_deflateEnd(zstream); kfree(zstream->workspace); + + if (c->tmp) + free_page((unsigned long)c->tmp); } static void err_compression_marker(struct drm_i915_error_state_buf *m) @@ -245,28 +263,34 @@ static void err_compression_marker(struct drm_i915_error_state_buf *m) #else -static bool compress_init(struct z_stream_s *zstream) +struct compress { +}; + +static bool compress_init(struct compress *c) { return true; } -static int compress_page(struct z_stream_s *zstream, +static int compress_page(struct compress *c, void *src, struct drm_i915_error_object *dst) { unsigned long page; + void *ptr; page = __get_free_page(GFP_ATOMIC | __GFP_NOWARN); if (!page) return -ENOMEM; - dst->pages[dst->page_count++] = - memcpy((void *)page, src, PAGE_SIZE); + ptr = (void *)page; + if (!i915_memcpy_from_wc(ptr, src, PAGE_SIZE)) + memcpy(ptr, src, PAGE_SIZE); + dst->pages[dst->page_count++] = ptr; return 0; } -static void compress_fini(struct z_stream_s *zstream, +static void compress_fini(struct compress *c, struct drm_i915_error_object *dst) { } @@ -784,7 +808,7 @@ i915_error_object_create(struct drm_i915_private *i915, struct i915_ggtt *ggtt = &i915->ggtt; const u64 slot = ggtt->error_capture.start; struct drm_i915_error_object *dst; - struct z_stream_s zstream; + struct compress compress; unsigned long num_pages; struct sgt_iter iter; dma_addr_t dma; @@ -804,7 +828,7 @@ i915_error_object_create(struct drm_i915_private *i915, dst->page_count = 0; dst->unused = 0; - if (!compress_init(&zstream)) { + if (!compress_init(&compress)) { kfree(dst); return NULL; } @@ -817,7 +841,7 @@ i915_error_object_create(struct drm_i915_private *i915, I915_CACHE_NONE, 0); s = io_mapping_map_atomic_wc(&ggtt->mappable, slot); - ret = compress_page(&zstream, (void __force *)s, dst); + ret = compress_page(&compress, (void __force *)s, dst); io_mapping_unmap_atomic(s); if (ret) @@ -832,7 +856,7 @@ unwind: dst = NULL; out: - compress_fini(&zstream, dst); + compress_fini(&compress, dst); ggtt->base.clear_range(&ggtt->base, slot, PAGE_SIZE); return dst; } -- 2.20.1