linux-kernel - Re: [PATCH -mm] timer: parenthesis fix in tbase_get

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Thu, 10 May 2007 13:56:27 +0530
From:	"Satyam Sharma" <satyam.sharma@...il.com>
To:	"Pallipadi, Venkatesh" <venkatesh.pallipadi@...el.com>
Cc:	"Jarek Poplawski" <jarkao2@...pl>,
	"Andrew Morton" <akpm@...ux-foundation.org>,
	linux-kernel@...r.kernel.org, "Oleg Nesterov" <oleg@...sign.ru>
Subject: Re: [PATCH -mm] timer: parenthesis fix in tbase_get_deferrable() etc.

> >> >> >  static inline unsigned int
> >tbase_get_deferrable(tvec_base_t *base)
> >> >> >  {
> >> >> > -  return ((unsigned int)(unsigned long)base &
> >> >TBASE_DEFERRABLE_FLAG);
> >> >> > +  return (unsigned int)((unsigned long)base &
> >> >TBASE_DEFERRABLE_FLAG);
> >> >> >  }
> >> >...
> >> >> The change makes sense, but does it actually "fix" anything?
> >> >>
> >> >
> >> >Yes - this first place fixes logical error, so it's a sin
> >> >- even if not punishable in practice. (It's also unnecessary
> >> >test for long to int conversion.)
> >> >
> >>
> >> I am sorry, I don't understand. What is the logical error in
> >the first
> >> one?
> >>
> >> Actually, your change makes it different from what was originally
> >> indended.
> >> Original intention was to type convert base to a 32 bit value and
> >> bitwise& with FLAG.
> >
> >But that is not what the original code is doing. If you wanted to
> >typecast "base" to "a 32 bit value" then you should've used u32
> >instead.
> >
> >Anyway, if you originally intended to actually typecast "base" to
> >unsigned int, then you could do that directly without typecasting it
> >first to unsigned long (unnecessarily) and then to unsigned int. Of
> >course, if your system implements a pointer as something bigger than
> >unsigned int (which is what you eventually convert "base" to), then
> >you're screwed anyway and the intermediate typecast to unsigned long
> >doesn't buy you anything at all.
>
> On a 64 bit system, converting pointer to int causes unnecessary
> compiler
> warning, and intermediate long conversion was to avoid that. I will have

Whoa! Hello, hold on, just wait a second there. Do you _really_ want
an unsigned int return out of tbase_get_deferrable() or will an
unsigned long do? If the rest of your code is fine with unsigned long,
then I'd suggest something like:

static inline unsigned long tbase_get_deferrable(tvec_base_t *base)
{
	return ((unsigned long)base & TBASE_DEFERRABLE_FLAG);
}

I don't really know your code (so I could be horribly incorrect here),
but personally I would prefer *heeding* to that warning than _hiding_
it -- it's not unnecessary, it's telling you that you're *losing* data
by converting a pointer (which is always unsigned long) to unsigned
int for 64-bit platforms where sizeof(void *) == sizeof(unsigned long)
== 8 bytes, but sizeof(unsigned int) == 4.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/