[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20260121095405.58fa513f@pumpkin>
Date: Wed, 21 Jan 2026 09:54:05 +0000
From: David Laight <david.laight.linux@...il.com>
To: David Desobry <david.desobry@...malgen.com>
Cc: "H. Peter Anvin" <hpa@...or.com>, tglx@...nel.org, mingo@...hat.com,
bp@...en8.de, dave.hansen@...ux.intel.com, x86@...nel.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] x86/lib: Optimize num_digits() and fix INT_MIN
overflow
On Wed, 21 Jan 2026 01:04:39 +0100
David Desobry <david.desobry@...malgen.com> wrote:
> On 1/21/26 00:49, H. Peter Anvin wrote:
..
> Actually, the V3 change:
> if (val < 0) {
> - d++;
> - val = -val;
> + d = 1;
> + v = -val;
> + } else {
> + d = 0;
> + v = val;
> }
> reintroduced the undefined behavior for val == INT_MIN.
> So this V3 version is incorrect.
Change to:
v = -(val + 1); v++;
The compiler notices the two '+ 1' cancel each other out.
David
> I'm not familiar enough with the rest of the codebase to know if
> changing the function signature to unsigned int is correct here.
In theory you'd need to change the name and all the callers.
David
Powered by blists - more mailing lists