On Tue, 18 Mar 2014, Joe Perches wrote:
> On Tue, 2014-03-18 at 13:55 +0100, Geert Uytterhoeven wrote:
> >
> > #define dprintk(flg, fmt, ...) no_printk(fmt, ##__VA_ARGS__)
>
> Fine, but with a correction.
>
> no_printk keeps all side effects like performing any function calls made
> by the statement or accessing any volatiles.
>
> Using
>
> do { if (0) no_printk(fmt, ##__VA_ARGS__); } while (0)
>
> does not have any side-effects.
I take your point about having the compiler check arguments when
NDEBUG & flg == 0.
As for side-effects, chip register accesses would be affected if dprintk()
expanded to no_printk() when NDEBUG & flg == 0.
E.g. NCR5380.c line 1213:
dprintk(NDEBUG_INTR, "scsi : unknown interrupt, BASR 0x%X, MR 0x%X, SR 0x%x\n",
basr, NCR5380_read(MODE_REG), NCR5380_read(STATUS_REG));
I don't want to re-introduce side-effects into a dozen different NCR5380
drivers on three different architectures when I can test only one of those
drivers. It's difficult to get good code coverage even for one driver.
--
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
