Add the list and Khem in case anyone is interested in Richard's comments
below.
../Randy
On 2022-12-05 08:47, Richard Purdie wrote:
Hi Yash,
On Mon, 2022-12-05 at 13:06 +0000, Shinde, Yash wrote:
PFA document for GCC patch analysis. There were about 28 patches in
GCC sources out of which 3 patches are with the upstream status of
‘Submitted’ and other 10 are ‘Pending’. I had analysed the current
status of these patches by checking them with the GCC patch list and
archive list and documented the same in the attached document.
Out of the 3 patches with ‘submitted’ as the status,
- The 1st patch is submitted but changes are not reflected in the
upstream. (hardcoded-paths.patch)
We should remind them about this one.
- The 2nd patch is not found in the upstream (0014-Don-t-search-host-
directory-during-relink-if-inst_pr.patch), was not found in patch
list or archive list and
This one is tricky since it is a libtool change. Was the change
accepted by libtool? If it was, we might be able to then persuade gcc
that they need it. It would mean updating gcc's libtool and I don't
know how they do that though.
- For the 3rd patch 1 hunk taken to upstream and can be marked as
'backported' and the other hunk is not taken to upstream yet.
(prefix-map-realpath.patch)
All the patch links are given in the doc.
I thought both pieces of this had merged?
https://gcc.gnu.org/git/?p=gcc.git;a=commitdiff;h=11543b27fe16d81ca5483ecb98ec7a5b2426e0c0
https://gcc.gnu.org/git/?p=gcc.git;a=commitdiff;h=e5c15eb183f17e806ad6b58c9497321ded87866f
We should update the patch to mention it is now a Backport with those
two links?
For the patches with ‘pending’ as the status, they are pending from
quite some time. Should we contact the author of the patch or the
community for these patches?
Please guide me on how to proceed further in GCC patch analysis and
what are the next steps.
For the Pending patches I'll add my comments below.
0013-Ensure-target-gcc-headers-can-be-included.patch
----------------------------------------------------
These headers are being installed under a prefix in the sysroot. We
need to check if that prefix is as a result of our own changes or
something upstream does. If upsptream uses the prefix, we should talk
to them about whether we can expect sysroot layout like this to work.
0005-Use-the-defaults.h-in-B-instead-of-S-and-t-oe-in-B.patch
-------------------------------------------------------------
I suspect this is inappropriate for upstream and should be marked as such.
0004-Pass-CXXFLAGS_FOR_BUILD-in-a-couple-of-places-to-avo.patch
---------------------------------------------------------------
Looks like it could be submitted/discussed upstream
0002-gcc-poison-system-directories.patch
----------------------------------------
Mark inappropriate?
0009-gcc-armv4-pass-fix-v4bx-to-linker-to-support-EABI.patch
------------------------------------------------------------
Looks like it could be submitted/discussed upstream
0017-Search-target-sysroot-gcc-version-specific-dirs-with.patch
---------------------------------------------------------------
Looks like it should be discussed upstream?
0008-libtool.patch
------------------
Would probably need to go to libtool first.
0020-libgcc_s-Use-alias-for-__cpu_indicator_init-instead-.patch
---------------------------------------------------------------
Check whether we still need it? (i686 musl?)
0018-Add-ssp_nonshared-to-link-commandline-for-musl-targe.patch
---------------------------------------------------------------
Check whether we still need it? (musl related?)
0006-cpp-honor-sysroot.patch
----------------------------
reproduce original issue, if still present submit upsteam, or drop
So in summary, there are a few issues we probably reproduce and discuss
with upstream.
Cheers,
Richard
--
# Randy MacLeod
# Wind River Linux
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#174293):
https://lists.openembedded.org/g/openembedded-core/message/174293
Mute This Topic: https://lists.openembedded.org/mt/95468728/21656
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-
