Fix strange panic message selection logic when swiotlb fills up

PATCH RESEND In review Casey Dahlin 2 participants 2 messages Aug 20, 2009 Library code

Share & tools

Casey Dahlin

author

2009-08-20 19:51 UTC #0 lore raw

PATCH RESEND 1/1 Fix strange panic message selection logic when swiotlb fills up

swiotlb_full in lib/swiotlb.c throws one of two panic messages based on whether the direction of transfer is from the device or to the device. The logic around this is somewhat weird in the case of bidirectional transfers. It appears to want to throw both in succession, but since its a panic only the first makes it.

This patch adds a third, separate error for DMA_BIDIRECTIONAL to make things a bit clearer.

Signed-off-by Casey Dahlin <[email protected]>

---
 lib/swiotlb.c |   15 +++++++++------
 1 files changed, 9 insertions(+), 6 deletions(-)

lib/swiotlb.c +9 -6

index bffe6d7..35e01b3 100644
@@ -625,12 +625,15 @@ swiotlb_full(struct device *dev, size_t size, int dir, int do_panic)
 	printk(KERN_ERR "DMA: Out of SW-IOMMU space for %zu bytes at "
 	       "device %s\n", size, dev ? dev_name(dev) : "?");
 
-	if (size > io_tlb_overflow && do_panic) {
-		if (dir == DMA_FROM_DEVICE || dir == DMA_BIDIRECTIONAL)
-			panic("DMA: Memory would be corrupted\n");
-		if (dir == DMA_TO_DEVICE || dir == DMA_BIDIRECTIONAL)
-			panic("DMA: Random memory would be DMAed\n");
-	}
+	if (size <= io_tlb_overflow || !do_panic)
+		return;
+
+	if (dir == DMA_BIDIRECTIONAL)
+		panic("DMA: Random memory could be corrupted or DMAed\n");
+	if (dir == DMA_FROM_DEVICE)
+		panic("DMA: Memory would be corrupted\n");
+	if (dir == DMA_TO_DEVICE)
+		panic("DMA: Random memory would be DMAed\n");
 }
 
 /*

Andrew Morton

maintainer

2009-08-20 23:25 UTC #1 lore raw

Quoted 29 lines · On Thu, 20 Aug 2009 15:52:28 -0400 Casey Dahlin <[email protected]> wrote:

swiotlb_full in lib/swiotlb.c throws one of two panic messages based on whether the direction of transfer is from the device or to the device. The logic around this is somewhat weird in the case of bidirectional transfers. It appears to want to throw both in succession, but since its a panic only the first makes it.
This patch adds a third, separate error for DMA_BIDIRECTIONAL to make things a bit clearer.
Signed-off-by Casey Dahlin <[email protected]>
---
 lib/swiotlb.c |   15 +++++++++------
 1 files changed, 9 insertions(+), 6 deletions(-)
lib/swiotlb.c +1 -6
index bffe6d7..35e01b3 100644
@@ -625,12 +625,15 @@ swiotlb_full(struct device *dev, size_t size, int dir, int do_panic)
 	printk(KERN_ERR "DMA: Out of SW-IOMMU space for %zu bytes at "
 	       "device %s\n", size, dev ? dev_name(dev) : "?");
 
-	if (size > io_tlb_overflow && do_panic) {
-		if (dir == DMA_FROM_DEVICE || dir == DMA_BIDIRECTIONAL)
-			panic("DMA: Memory would be corrupted\n");
-		if (dir == DMA_TO_DEVICE || dir == DMA_BIDIRECTIONAL)
-			panic("DMA: Random memory would be DMAed\n");
-	}
+	if (size <= io_tlb_overflow || !do_panic)

lib/swiotlb.c swiotlb_full(struct device *dev, size_t size, int dir, int do_panic) · line 638

Quoted 10 lines

lib/swiotlb.c +8 -0

+		return;
+
+	if (dir == DMA_BIDIRECTIONAL)
+		panic("DMA: Random memory could be corrupted or DMAed\n");
+	if (dir == DMA_FROM_DEVICE)
+		panic("DMA: Memory would be corrupted\n");
+	if (dir == DMA_TO_DEVICE)
+		panic("DMA: Random memory would be DMAed\n");
 }

Seems sane to me.

While we're there, how about we make the final message less ambiguous?

lib/swiotlb.c renamed from lib/swiotlb.c~lib-swiotlbc-fix-strange-panic-message-selection-logic-when-swiotlb-fills-up-fix +1 -1

@@ -633,7 +633,7 @@ swiotlb_full(struct device *dev, size_t 
 	if (dir == DMA_FROM_DEVICE)
 		panic("DMA: Memory would be corrupted\n");
 	if (dir == DMA_TO_DEVICE)
-		panic("DMA: Random memory would be DMAed\n");
+		panic("DMA: Random memory would be DMAed from\n");
 }
 
 /*