On Oct 26, 2010, at 6:10 PM, Joe Perches <joe@xxxxxxxxxxx> wrote: > On Tue, 2010-10-26 at 18:00 -0700, Dmitry Torokhov wrote: >> On Oct 26, 2010, at 5:46 PM, Joe Perches <joe@xxxxxxxxxxx> wrote: >>> On Tue, 2010-10-26 at 17:40 -0700, Dmitry Torokhov wrote: >>>>> Mind if I move the time to the front like this? >>>>> +#define dbg(format, arg...) \ >>>>> +do { \ >>>>> + if (i8042_debug) \ >>>>> + printk(KERN_DEBUG KBUILD_MODNAME ": [%d]" format), \ >>>>> + (int)(jiffies - i8042_start_time), ##arg); \ >>>>> +} while (0) >>>> Another option would be to leave '\n' in dbg() and leave most of the >>>> callers as is. >>> That's really error prone though. >> In what way? > > $ grep -w dbg drivers/input/serio/i8042.c | cat -n > 1 dbg("%02x <- i8042 (flush, %s)", data, > 2 dbg("%02x -> i8042 (command)", command & 0xff); > 3 dbg("%02x -> i8042 (parameter)", param[i]); > 4 dbg(" -- i8042 (timeout)"); > 5 dbg(" -- i8042 (auxerr)"); > 6 dbg("%02x <- i8042 (return)", param[i]); > 7 dbg("%02x -> i8042 (kbd-data)", c); > 8 dbg("Extra keyboard ACK - filtered out\n"); > 9 dbg("Filtered out by platform filter\n"); > 10 if (irq) dbg("Interrupt %d, without any data", irq); > 11 dbg("MUX error, status is %02x, data is %02x", str, data); > 12 dbg("%02x <- i8042 (interrupt, %d, %d%s%s)", > 13 dbg("%02x <- i8042 (aux_test_irq, %s)", > 14 dbg(" -- i8042 (aux irq test timeout)"); > 15 dbg("%02x -> i8042 (panic blink)", 0xed); > 16 dbg("%02x -> i8042 (panic blink)", led); > > I think 2 defects in 16 uses is a pretty high error rate. But how not adding /n is it much better than having it if some later change will forget to add it to the message? -- Dmitry -- To unsubscribe from this list: send the line "unsubscribe linux-input" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html