This is the mail archive of the
mailing list for the glibc project.
Re: [PATCH][BZ #16214] Fix TLS access on S390 with -march=z10
- From: Andreas Krebbel <krebbel at linux dot vnet dot ibm dot com>
- To: Andreas Schwab <schwab at suse dot de>
- Cc: Siddhesh Poyarekar <siddhesh at redhat dot com>, libc-alpha at sourceware dot org, carlos at redhat dot com
- Date: Thu, 28 Nov 2013 12:22:39 +0100
- Subject: Re: [PATCH][BZ #16214] Fix TLS access on S390 with -march=z10
- Authentication-results: sourceware.org; auth=none
- References: <20131125055009 dot GK19834 at spoyarek dot pnq dot redhat dot com> <20131127210146 dot GA22879 at bart> <20131128043516 dot GD20495 at spoyarek dot pnq dot redhat dot com> <52971AFE dot 2070502 at linux dot vnet dot ibm dot com> <mvm4n6wdc0j dot fsf at hawking dot suse dot de>
On 28/11/13 12:03, Andreas Schwab wrote:
> Andreas Krebbel <firstname.lastname@example.org> writes:
>> - Adding a new symbol for __tls_get_addr and putting the GLIBC_PRIVATE version on it. The new
>> symbol can only be exported when removing the hidden attribute from the original function but this
>> led to the original function to be exported as __tls_get_addr@@GLIBC_2.3 what I'm trying to prevent.
> Does it work to use default_symbol_version with GLIBC_PRIVATE?
No. The hidden attribute is inherited as well. Since versioned_symbol expands to
default_symbol_version it would have been surprising if it would behave differently.