In dma_common_find_pages(), area->flags are compared directly with
VM_DMA_COHERENT. This works because VM_DMA_COHERENT is the only set
flag.

During development of a new feature (ASI [1]), a new VM flag is
introduced, and that flag can be injected into VM_DMA_COHERENT mappings
(among others).  The presence of that flag caused
dma_common_find_pages() to return NULL for VM_DMA_COHERENT addresses,
leading to a lot of problems ending in crashing during boot. It took a
bit of time to figure this problem out.

It was a mistake to inject a VM flag to begin with, but it took a
significant amount of debugging to figure out the problem. Most users of
area->flags use bitmasking rather than equivalency to check for flags.
Update dma_common_find_pages() and dma_common_free_remap() to do the
same, which would have avoided the boot crashing. Instead, add a warning
in dma_common_find_pages() if any extra VM flags are set to catch such
problems more easily during development.

No functional change intended.

[1]https://lore.kernel.org/lkml/20240712-asi-rfc-24-v1-0-144b319a40d8@google.com/

Signed-off-by: Yosry Ahmed <yosryahmed@google.com>
---
 kernel/dma/remap.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/kernel/dma/remap.c b/kernel/dma/remap.c
index 27596f3b4aef3..9e2afad1c6152 100644
--- a/kernel/dma/remap.c
+++ b/kernel/dma/remap.c
@@ -10,8 +10,10 @@ struct page **dma_common_find_pages(void *cpu_addr)
 {
 	struct vm_struct *area = find_vm_area(cpu_addr);
 
-	if (!area || area->flags != VM_DMA_COHERENT)
+	if (!area || !(area->flags & VM_DMA_COHERENT))
 		return NULL;
+	WARN(area->flags != VM_DMA_COHERENT,
+	     "unexpected flags in area: %p\n", cpu_addr);
 	return area->pages;
 }
 
@@ -61,7 +63,7 @@ void dma_common_free_remap(void *cpu_addr, size_t size)
 {
 	struct vm_struct *area = find_vm_area(cpu_addr);
 
-	if (!area || area->flags != VM_DMA_COHERENT) {
+	if (!area || !(area->flags & VM_DMA_COHERENT)) {
 		WARN(1, "trying to free invalid coherent area: %p\n", cpu_addr);
 		return;
 	}
-- 
2.46.0.76.ge559c4bf1a-goog

Thanks Yosry, this looks good to me.  I don't see a real urgency here,
so I plan to add it to the dma-mapping tree for 6.12.

On Mon, Aug 12, 2024 at 1:32 AM Christoph Hellwig <hch@lst.de> wrote:
>
> Thanks Yosry, this looks good to me.  I don't see a real urgency here,
> so I plan to add it to the dma-mapping tree for 6.12.
>

Yeah no urgency here, thanks for the swift response, Cristoph!