This is the mail archive of the
mailing list for the libc-ports project.
Re: [PATCH] AArch64 - Add parameter to __mount routine
- From: Venkataramanan Kumar <venkataramanan dot kumar at linaro dot org>
- To: Marcus Shawcroft <marcus dot shawcroft at gmail dot com>
- Cc: libc-ports at sourceware dot org, Patch Tracking <patch at linaro dot org>
- Date: Mon, 16 Sep 2013 19:05:13 +0530
- Subject: Re: [PATCH] AArch64 - Add parameter to __mount routine
- Authentication-results: sourceware.org; auth=none
- References: <CAJK_mQ2OQPP7_RFcMJc2NP7Wz7hX2AHQ9aVb8-uKi+MZN4UFsA at mail dot gmail dot com> <CAFqB+PwpaUernHtV2acN03D=vhp7C09Vx0E5jVw9Ek=5PJ2MgA at mail dot gmail dot com>
On 16 September 2013 17:01, Marcus Shawcroft <email@example.com> wrote:
> On 15 September 2013 13:45, Venkataramanan Kumar
> <firstname.lastname@example.org> wrote:
>> Hi Maintainers,
>> Attached is a small patch which adds a parameter to __mcount routine.
>> We would like to make the compiler pass the "frompc" address from the function
>> that calls __mcount during profiling.
> This is an ABI change, however, given that there is no functioning
> profiling solution using the existing interface I don't think that
> such a change is a problem. Does anyone object?
>> ref: http://gcc.gnu.org/ml/gcc-patches/2013-08/msg01556.html
>> This patch is cross tested in ARMV8 foundation model. No new regressions found.
>> If it is ok, someone please commit on my behalf.
> No regressions is a good start. How did you test the the combined
> gcc/glibc patches for profiling are working correctly?
I am cross compiling some simple tests cases and running them under in
ARMV8 model by invoking the runtime linker.
Basically do functional testing for these test cases by checking the
available gprof options.