Skip to contentSkip to conversation

libahci: Fix bug in storing EM messages

PATCH Applied Harry Zhang 3 participants 3 messages Jun 24, 2010 – Jul 1, 2010 libata

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 libahci: Fix bug in storing EM messages

In function ahci_store_em_buffer(), if the input (signed char*) buffer contains negative data, the constructed 32-bit long message data may be wrong.

---
 drivers/ata/libahci.c |    5 +++--
 1 files changed, 3 insertions(+), 2 deletions(-)
drivers/ata/libahci.c +3 -2
index 1984a6e..7b273f0 100644
@@ -324,6 +324,7 @@ static ssize_t ahci_store_em_buffer(struct device *dev,
 	struct ahci_host_priv *hpriv = ap->host->private_data;
 	void __iomem *mmio = hpriv->mmio;
 	void __iomem *em_mmio = mmio + hpriv->em_loc;
+	const unsigned char *msg_buf = buf;
 	u32 em_ctl, msg;
 	unsigned long flags;
 	int i;
@@ -343,8 +344,8 @@ static ssize_t ahci_store_em_buffer(struct device *dev,
 	}
 
 	for (i = 0; i < size; i += 4) {
-		msg = buf[i] | buf[i + 1] << 8 |
-		      buf[i + 2] << 16 | buf[i + 3] << 24;
+		msg = msg_buf[i] | msg_buf[i + 1] << 8 |
+		      msg_buf[i + 2] << 16 | msg_buf[i + 3] << 24;
 		writel(msg, em_mmio + i);
 	}
 
Signature · 1 line
1.6.4.2
#1 lore raw
Quoted 3 lines · Harry Zhang

In function ahci_store_em_buffer(), if the input (signed char*) buffer contains negative data, the constructed 32-bit long message data may be wrong.

Explaining why it was wrong would be nice. Also is adding a comment explaining why the uchar pointer is needed. Nothing fancy, just simple /* avoid sign extension */ should do.

Other than that, Acked-by: Tejun Heo <[email protected]>

Thanks.

Signature · 1 line
tejun
Applied
#2 lore raw
In reply to Harry Zhang
Quoted 8 lines · Harry Zhang

In function ahci_store_em_buffer(), if the input (signed char*) buffer contains negative data, the constructed 32-bit long message data may be wrong.

---
  drivers/ata/libahci.c |    5 +++--
  1 files changed, 3 insertions(+), 2 deletions(-)

applied