This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH 2/2] Initialize tunable list with the GLIBC_TUNABLES environment variable
- From: "Carlos O'Donell" <carlos at redhat dot com>
- To: Andreas Schwab <schwab at suse dot de>, Siddhesh Poyarekar <sid at reserved-bit dot com>
- Cc: libc-alpha at sourceware dot org, roland at hack dot frob dot com, "Paul E. Murphy" <murphyp at linux dot vnet dot ibm dot com>, Andi Kleen <andi at firstfloor dot org>
- Date: Wed, 13 Jan 2016 10:37:05 -0500
- Subject: Re: [PATCH 2/2] Initialize tunable list with the GLIBC_TUNABLES environment variable
- Authentication-results: sourceware.org; auth=none
- References: <20160111111719 dot GA4183 at devel dot intra dot reserved-bit dot com> <mvmd1t81axj dot fsf at hawking dot suse dot de> <20160111144537 dot GA3334 at devel dot intra dot reserved-bit dot com> <5695BA23 dot 6080307 at redhat dot com> <20160113032725 dot GJ3334 at devel dot intra dot reserved-bit dot com> <mvm37u1x4a2 dot fsf at hawking dot suse dot de>
On 01/13/2016 03:38 AM, Andreas Schwab wrote:
> Siddhesh Poyarekar <sid@reserved-bit.com> writes:
>
>> A single load (and constructor call) does not matter since there's
>> nothing to race against. The case of concurrent constructor calls I
>> mentioned in the following paragraph; the load lock is sufficient to
>> protect it.
>
> PR19448 is about removing that lock.
If the code has comments that it needs protection from load lock then
such code will be reviewed as we fix BZ19448. Which is why I suggested
comments to that extent. The removal of the load lock won't be easy
and will require quite a bit of review.
Cheers,
Carlos.