From: Rick Edgecombe rick.p.edgecombe@intel.com
[ Upstream commit b9fa16949d18e06bdf728a560f5c8af56d2bdcaf ]
On TDX it is possible for the untrusted host to cause set_memory_encrypted() or set_memory_decrypted() to fail such that an error is returned and the resulting memory is shared. Callers need to take care to handle these errors to avoid returning decrypted (shared) memory to the page allocator, which could lead to functional or security issues.
DMA could free decrypted/shared pages if dma_set_decrypted() fails. This should be a rare case. Just leak the pages in this case instead of freeing them.
Signed-off-by: Rick Edgecombe rick.p.edgecombe@intel.com Signed-off-by: Christoph Hellwig hch@lst.de Signed-off-by: Sasha Levin sashal@kernel.org --- kernel/dma/direct.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)
diff --git a/kernel/dma/direct.c b/kernel/dma/direct.c index 98b2e192fd696..4d543b1e9d577 100644 --- a/kernel/dma/direct.c +++ b/kernel/dma/direct.c @@ -286,7 +286,7 @@ void *dma_direct_alloc(struct device *dev, size_t size, } else { ret = page_address(page); if (dma_set_decrypted(dev, ret, size)) - goto out_free_pages; + goto out_leak_pages; }
memset(ret, 0, size); @@ -307,6 +307,8 @@ void *dma_direct_alloc(struct device *dev, size_t size, out_free_pages: __dma_direct_free_pages(dev, page, size); return NULL; +out_leak_pages: + return NULL; }
void dma_direct_free(struct device *dev, size_t size, @@ -367,12 +369,11 @@ struct page *dma_direct_alloc_pages(struct device *dev, size_t size,
ret = page_address(page); if (dma_set_decrypted(dev, ret, size)) - goto out_free_pages; + goto out_leak_pages; memset(ret, 0, size); *dma_handle = phys_to_dma_direct(dev, page_to_phys(page)); return page; -out_free_pages: - __dma_direct_free_pages(dev, page, size); +out_leak_pages: return NULL; }