Re: [PATCH] i386 spinlocks should use the full 32 bits, not only 8 bits


On Tue, 11 Oct 2005, Eric Dumazet wrote:
>
> As NR_CPUS might be > 128, and every spining CPU decrements the lock, we need
> to use more than 8 bits for a spinlock. The current (i386/x86_64)
> implementations have a (theorical) bug in this area.

I don't think there are any x86 machines with > 128 CPU's right now.

The advantage of the byte lock is that a "movb $0" is three bytes shorter 
than a "movl $0". And that's the unlock sequence.

		Linus
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Follow-Ups:
- Re: [PATCH] i386 spinlocks should use the full 32 bits, not only 8 bits
  - From: Andi Kleen <[email protected]>
- Re: [PATCH] i386 spinlocks should use the full 32 bits, not only 8 bits
  - From: Eric Dumazet <[email protected]>

References:
- i386 spinlock fairness: bizarre test results
  - From: Chuck Ebbert <[email protected]>
- Re: i386 spinlock fairness: bizarre test results
  - From: Alan Cox <[email protected]>
- Re: i386 spinlock fairness: bizarre test results
  - From: Linus Torvalds <[email protected]>
- [PATCH] i386 spinlocks should use the full 32 bits, not only 8 bits
  - From: Eric Dumazet <[email protected]>

Prev by Date: Re: 2.6.14-rc4-rt1
Next by Date: 2.6.14-rc4-rt1 - scsi_eh_4/7835[CPU#0]: BUG in __schedule at kernel/sched.c:3326
Previous by thread: [PATCH] i386 spinlocks should use the full 32 bits, not only 8 bits
Next by thread: Re: [PATCH] i386 spinlocks should use the full 32 bits, not only 8 bits
Index(es):
- Date
- Thread

[Index of Archives] [Kernel Newbies] [Netfilter] [Bugtraq] [Photo] [Stuff] [Gimp] [Yosemite News] [MIPS Linux] [ARM Linux] [Linux Security] [Linux RAID] [Video 4 Linux] [Linux for the blind] [Linux Resources]