Re: [PATCH] riscv/atomic: Do proper sign extension also for unsigned in arch_cmpxchg

From: Andreas Schwab <schwab@suse.de>
To: David Laight <david.laight.linux@gmail.com>
Cc: linux-riscv@lists.infradead.org,  linux-kernel@vger.kernel.org
Subject: Re: [PATCH] riscv/atomic: Do proper sign extension also for unsigned in arch_cmpxchg
Date: Mon, 03 Feb 2025 09:42:02 +0100	[thread overview]
Message-ID: <mvmseov2zdx.fsf@suse.de> (raw)
In-Reply-To: <20250130213504.285b4675@pumpkin> (David Laight's message of "Thu, 30 Jan 2025 21:35:04 +0000")

On Jan 30 2025, David Laight wrote:

> On Thu, 30 Jan 2025 10:25:38 +0100
> Andreas Schwab <schwab@suse.de> wrote:
>
>> Sign extend also an unsigned compare value to match what lr.w is doing.
>> Otherwise try_cmpxchg may spuriously return true when used on a u32 value
>> that has the sign bit set, as it happens often in inode_set_ctime_current.
>> 
>> Do this in three conversion steps.  The first conversion to long is needed
>> to avoid a -Wpointer-to-int-cast warning when arch_cmpxchg is used with a
>> pointer type.
>
> Doesn't that break things by discarding the high bits of a pointer value?

If you have 32-bit pointers then the conversions are no-ops.

-- 
Andreas Schwab, SUSE Labs, schwab@suse.de
GPG Key fingerprint = 0196 BAD8 1CE9 1970 F4BE  1748 E4D4 88E3 0EEA B9D7
"And now for something completely different."

_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv