This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |
Other format: | [Raw text] |
On 06/13/2016 04:39 PM, Joseph Myers wrote:
Sorry for the delay. I had to wait for some test machines to be prepared. I have posted a patchset where the common libc_ifunc macro is adjusted and used by all architectures including s390.On Mon, 13 Jun 2016, Stefan Liebler wrote:On 06/10/2016 06:59 PM, Joseph Myers wrote:Are the debug info issues you describe unique to S/390 because of some peculiarity of how S/390, and only S/390, defines IFUNC resolvers in glibc, or do they apply to any other architectures as well and so need fixes there?The libc_ifunc macro in include/libc-symbols.h also uses the asm-name and the debug-information contains the DW_AT_linkage_name field.Then I think the goal should be to arrange things so that as much infrastructure as possible is shared between architectures defining IFUNC resolvers in C code, so that they all avoid this pitfall, rather than just fixing one architecture.
Thus this patch is superseded by the patch series:"[PATCH 1/8] Use gcc attribute ifunc in libc_ifunc macro instead of inline assembly due to false debuginfo."
https://www.sourceware.org/ml/libc-alpha/2016-06/msg00896.html
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |