Skip to contentSkip to conversation

dma-pool: warn when coherent pool is depleted

PATCH Guenter Roeck 3 participants 3 messages Jun 21, 2020 – Jun 29, 2020

lore
Share & tools
lore

Guenter Roeck

author Tested-by
#0 lore raw
Quoted 10 lines · David Rientjes

When a DMA coherent pool is depleted, allocation failures may or may not get reported in the kernel log depending on the allocator.

The admin does have a workaround, however, by using coherent_pool= on the kernel command line.

Provide some guidance on the failure and a recommended minimum size for the pools (double the size).

Also confirmed that coherent_pool=256k works around the crash I had observed.

Guenter

Quoted 15 lines
---
 kernel/dma/pool.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
kernel/dma/pool.c +1 -1
@@ -239,12 +239,16 @@ void *dma_alloc_from_pool(struct device *dev, size_t size,
 	}
 
 	val = gen_pool_alloc(pool, size);
-	if (val) {
+	if (likely(val)) {
 		phys_addr_t phys = gen_pool_virt_to_phys(pool, val);
 

kernel/dma/pool.c void *dma_alloc_from_pool(struct device *dev, size_t size, · line 254

Quoted 10 lines
kernel/dma/pool.c +4 -0
 		*ret_page = pfn_to_page(__phys_to_pfn(phys));
 		ptr = (void *)val;
 		memset(ptr, 0, size);
+	} else {
+		WARN_ONCE(1, "DMA coherent pool depleted, increase size "
+			     "(recommended min coherent_pool=%zuK)\n",
+			  gen_pool_size(pool) >> 9);
 	}
 	if (gen_pool_avail(pool) < atomic_pool_size)
 		schedule_work(&atomic_pool_work);

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu

#1 lore raw
Quoted 9 lines · Guenter Roeck
Quoted 9 lines

When a DMA coherent pool is depleted, allocation failures may or may not get reported in the kernel log depending on the allocator.

The admin does have a workaround, however, by using coherent_pool= on the kernel command line.

Provide some guidance on the failure and a recommended minimum size for the pools (double the size).

Quoted 7 lines
Quoted 1 line

Also confirmed that coherent_pool=256k works around the crash I had observed.

Thanks Guenter. Christoph, does it make sense to apply this patch since there may not be an artifact left behind in the kernel log on allocation failure by the caller?

Quoted 28 lines

Guenter

Quoted 25 lines
---
 kernel/dma/pool.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
kernel/dma/pool.c +5 -1
@@ -239,12 +239,16 @@ void *dma_alloc_from_pool(struct device *dev, size_t size,
 	}
 
 	val = gen_pool_alloc(pool, size);
-	if (val) {
+	if (likely(val)) {
 		phys_addr_t phys = gen_pool_virt_to_phys(pool, val);
 
 		*ret_page = pfn_to_page(__phys_to_pfn(phys));
 		ptr = (void *)val;
 		memset(ptr, 0, size);
+	} else {
+		WARN_ONCE(1, "DMA coherent pool depleted, increase size "
+			     "(recommended min coherent_pool=%zuK)\n",
+			  gen_pool_size(pool) >> 9);
 	}
 	if (gen_pool_avail(pool) < atomic_pool_size)
 		schedule_work(&atomic_pool_work);