Re: [tip:core/locking] locking, x86: Slightly shorten __ticket_spin_trylock()
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
- Subject: Re: [tip:core/locking] locking, x86: Slightly shorten __ticket_spin_trylock()
- From: "Jan Beulich" <JBeulich@xxxxxxxxxx>
- Date: Wed, 02 Dec 2009 14:59:42 +0000
- Cc: <a.p.zijlstra@xxxxxxxxx>, "Ingo Molnar" <mingo@xxxxxxx>, <tglx@xxxxxxxxxxxxx>, <torvalds@xxxxxxxxxxxxxxxxxxxx>, <mingo@xxxxxxxxxx>, <npiggin@xxxxxxx>, <linux-kernel@xxxxxxxxxxxxxxx>, <linux-tip-commits@xxxxxxxxxxxxxxx>, <hpa@xxxxxxxxx>
- In-reply-to: <4B167B77.8030502@xxxxxxxxxx>
- References: <4B0FF9AC0200007800022713@xxxxxxxxxxxxxxxxxx> <tip-133ec7a235160dd44cbd4d82fff65a9983331df9@xxxxxxxxxxxxxx> <20091202132937.GA1564@xxxxxxx> <4B168293020000780002308E@xxxxxxxxxxxxxxxxxx> <4B1675CA.1020504@xxxxxxxxxx> <4B1686F702000078000230A6@xxxxxxxxxxxxxxxxxx> <4B167B77.8030502@xxxxxxxxxx>
>>> Avi Kivity <avi@xxxxxxxxxx> 02.12.09 15:36 >>>
>On 12/02/2009 04:25 PM, Jan Beulich wrote:
>>>>> Avi Kivity<avi@xxxxxxxxxx> 02.12.09 15:12>>>
>>>>>
>>> Wouldn't 'u8 ret', as an additional argument be sufficient? gcc still
>>> ought to be able to use the same register for new and ret if we remove
>>> the early clobber.
>>>
>> We can't (validly) remove the early-clobber, otherwise the compiler
>> would be permitted to use a register also used for addressing the lock
>> structure for "new".
>>
>
>I meant for 'ret'.
But unless you drop the early-clobber on new, the compiler will have
to allocate another register, which I think is not desirable here.
Jan
--
To unsubscribe from this list: send the line "unsubscribe linux-tip-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
[Index of Archives]
[Linux Stable Commits]
[Linux Stable Kernel]
[Linux Kernel]
[Linux USB Devel]
[Linux Video &Media]
[Linux Audio Users]
[Yosemite News]
[Linux SCSI]