diff mbox series

[2/2] drm/ttm: Cache dma pool decrypted pages when AMD SEV is active

Message ID 20190823070828.18112-2-thomas_os@shipmail.org (mailing list archive)
State New, archived
Headers show
Series [1/2] drm/ttm, drm/vmwgfx: Have TTM support AMD SEV encryption | expand

Commit Message

Thomas Hellström (Intel) Aug. 23, 2019, 7:08 a.m. UTC
From: Thomas Hellstrom <thellstrom@vmware.com>

The TTM dma pool allocates coherent pages for use with TTM. When SEV is
active, such allocations become very expensive since the linear kernel
map has to be changed to mark the pages decrypted. So to avoid too many
such allocations and frees, cache the decrypted pages even if they
are in the normal cpu caching state, where otherwise the pool frees them
immediately when unused.

Tested with vmwgfx on SEV-ES.

Cc: Christian König <christian.koenig@amd.com>
Cc: Thomas Lendacky <thomas.lendacky@amd.com>
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
---
 drivers/gpu/drm/ttm/ttm_page_alloc_dma.c | 19 ++++++++++++++-----
 1 file changed, 14 insertions(+), 5 deletions(-)
diff mbox series

Patch

diff --git a/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c b/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
index c7e223c4f26c..a4445a83bc96 100644
--- a/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
+++ b/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
@@ -999,7 +999,7 @@  void ttm_dma_unpopulate(struct ttm_dma_tt *ttm_dma, struct device *dev)
 	struct dma_pool *pool;
 	struct dma_page *d_page, *next;
 	enum pool_type type;
-	bool is_cached = false;
+	bool immediate_free = false;
 	unsigned count, i, npages = 0;
 	unsigned long irq_flags;
 
@@ -1034,8 +1034,17 @@  void ttm_dma_unpopulate(struct ttm_dma_tt *ttm_dma, struct device *dev)
 	if (!pool)
 		return;
 
-	is_cached = (ttm_dma_find_pool(pool->dev,
-		     ttm_to_type(ttm->page_flags, tt_cached)) == pool);
+	/*
+	 * If memory is cached and sev encryption is not active, allocating
+	 * and freeing coherent memory is relatively cheap, so we can free
+	 * it immediately. If sev encryption is active, allocating coherent
+	 * memory involves a call to set_memory_decrypted() which is very
+	 * expensive, so cache coherent pages is sev is active.
+	 */
+	immediate_free = (ttm_dma_find_pool
+			  (pool->dev,
+			   ttm_to_type(ttm->page_flags, tt_cached)) == pool &&
+			  !sev_active());
 
 	/* make sure pages array match list and count number of pages */
 	count = 0;
@@ -1050,13 +1059,13 @@  void ttm_dma_unpopulate(struct ttm_dma_tt *ttm_dma, struct device *dev)
 			d_page->vaddr &= ~VADDR_FLAG_UPDATED_COUNT;
 		}
 
-		if (is_cached)
+		if (immediate_free)
 			ttm_dma_page_put(pool, d_page);
 	}
 
 	spin_lock_irqsave(&pool->lock, irq_flags);
 	pool->npages_in_use -= count;
-	if (is_cached) {
+	if (immediate_free) {
 		pool->nfrees += count;
 	} else {
 		pool->npages_free += count;