Re: [PATCH] perf: fix bug in isupper() and islower()

[Arnaldo Carvalho de Melo <acme@ghostprotocols.net>]

Em Mon, Apr 01, 2013 at 03:19:28PM +0900, Namhyung Kim escreveu:
> On Fri, 29 Mar 2013 12:29:50 -0700, Sukadev Bhattiprolu wrote:
> > Subject: [PATCH] perf: fix bug in isupper() and islower()

> > One of the reasons 'perf test' is failing on Power appears to be due to
> > a bug in isupper().

> > isupper(c) and islower(c) should be checking 'c' against the mask 0x20.
> > Instead they are checking sane_ctype[c] which causes isupper() to be true
> > for lower case letters.

> Indeed!
> Acked-by: Namhyung Kim <namhyung@kernel.org>

But can we try to use the same defs as for the kernel? Or at least sync
this code against git.git, that is where it comes from?

There we have:

#define islower(x) sane_iscase(x, 1)
#define isupper(x) sane_iscase(x, 0)

#define sane_istest(x,mask) ((sane_ctype[(unsigned char)(x)] & (mask)) != 0)

static inline int sane_iscase(int x, int is_lower)
{
        if (!sane_istest(x, GIT_ALPHA))
                return 0;

        if (is_lower)
                return (x & 0x20) != 0;
        else
                return (x & 0x20) == 0;
}
----------------------------------------------------------------------------

In the kernel we have instead:

include/linux/ctype.h

#define _U      0x01    /* upper */
#define __ismask(x) (_ctype[(int)(unsigned char)(x)])
#define isupper(c)      ((__ismask(c)&(_U)) != 0)

----------------------------------------------------------------------------

I'm merging this fix now, as it improves things, but this seemingly
needless speciations looks insane, would be great to have at least
tools/perf/  using the same stuff as its landlord, the kernel.

- Arnaldo