Re: [PATCH 12/29] x86, tsx: Add a per thread transaction disablecount
From: Borislav Petkov
Date: Sat Mar 23 2013 - 13:32:53 EST
On Sat, Mar 23, 2013 at 10:16:28AM -0700, Linus Torvalds wrote:
> On Sat, Mar 23, 2013 at 8:52 AM, Borislav Petkov <bp@xxxxxxxxx> wrote:
> >
> > Really? Let's see:
>
> Your test seems to assume that a single bit is sufficient, which
> sounds unlikely. If you have any kind of nesting going on, you need
> more than one bit. Then add/sub isn't a single "xor" of a bit any
> more.
Right, but does nesting even apply here? I mean, AFAICR the commit
message, we're talking about per-thread transaction disabling. IOW, this
is an on/off button used in a boolean context.
Or is this going to be used as a mechanism to tisable TSX when max
nesting level has been reached?
Hmmm.
Btw, the disable/reenable_txn() aren't used anywhere in the patchset and
I'd actually like to get an idea where/how those are used.
Thanks.
--
Regards/Gruss,
Boris.
Sent from a fat crate under my desk. Formatting is fine.
--
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/