Borislav Petkov wrote:
On Fri, Jun 30, 2006 at 01:27:30AM +0900, Tejun Heo wrote:
<snip>
@@ -805,11 +807,16 @@ extern void ata_do_eh(struct ata_port *a
/*
* printk helpers
*/
-#define ata_port_printk(ap, lv, fmt, args...) \
- printk(lv"ata%u: "fmt, (ap)->id , ##args)
+#define ata_port_printk(ap, lv, fmt, args...) do { \
+ if (unlikely((ap)->msg_enable & (0xFF & (lv)))) \
^^^^why?
Because the decimal value of ATA_MSG_TRACE, for example, is 128 (hex 0x80) and
when i shift a 1 128 time to the left, the datatype, in our case an int, i
think (dunno what gcc does exactly here) is too small for the shift. Besides,
ap->msg_enable is an u32 so that if would never be true for something bigger
than 0x20.
Please make
+ ATA_MSG_ERR = 0,
+ ATA_MSG_WARN = 1,
+ ATA_MSG_DRV = 2,
+ ATA_MSG_INFO = 3, /* revalidation messages, EH progress */
+ ATA_MSG_VDEBUG = 4, /* verbose hot path */
+ ATA_MSG_CMD = 5, /* issue / completion */
+ ATA_MSG_SG = 6, /* SG map/unmap handling */
+ ATA_MSG_TRACE = 7, /* function tracing, e.g. enter/exit */
And test for msg_enable & (1 << (lv)). Otherwise the array will be
unnecessarily large.
--
tejun
-
: send the line "unsubscribe linux-ide" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
