Skip to content

Merge 20260807 - #258

Open
amd-shahab wants to merge 9 commits into
amd-stagingfrom
users/shvahedi/merge-20260807
Open

Merge 20260807#258
amd-shahab wants to merge 9 commits into
amd-stagingfrom
users/shvahedi/merge-20260807

Conversation

@amd-shahab

Copy link
Copy Markdown
Contributor

Motivation

Merge upstream master.

Technical Details

CI merge failed.

Test Plan

make check-gdb RUNTESTFLAGS="GDB_TESTCASE_OPTIONS='-O0'" TESTS="gdb.rocm/*.exp"

Test on GFX1201 and GFX1103.

Test Result

GFX1201

                === gdb Summary ===

# of expected passes            2522
# of expected failures          7
# of known failures             7
# of unsupported tests          1

GFX1103

FAIL: gdb.rocm/watchpoint-basic.exp: test_watchpoint_before_kernel: continue to watchpoint hit 1
FAIL: gdb.rocm/watchpoint-basic.exp: test_remove_watchpoint_inside_kernel: continue to watchpoint hit
FAIL: gdb.rocm/watchpoint-basic.exp: test_multiple_watchpoints: continue to watchpoint hit 1
FAIL: gdb.rocm/watchpoint-basic.exp: test_disable_enable_watchpoint: continue to watchpoint hit 1

                === gdb Summary ===

# of expected passes            2378
# of unexpected failures        4
# of expected failures          3
# of known failures             7
# of untested testcases         6
# of unsupported tests          9

watchpoint-basic failures are expected (AIROCGDB-555) and must be added as XFAIL.

hjl-tools and others added 9 commits August 6, 2026 09:32
The ELF backend gc_mark_extra_sections function may return false for
error and bfd_gc_sections may return false on invalid input:

ld: pr34448-bug_18.o: bad reloc symbol index (0xf2000005 >= 0x13) for offset 0x4 in section `.text.get_tls[get_tls]'

Change bfd_elf_gc_sections to return false if gc_mark_extra_sections
return false.  Change lang_gc_sections to check bfd_gc_sections return
and report the fatal error.

bfd/
	PR ld/34448
	* elflink.c (bfd_elf_gc_sections): Return false if
	gc_mark_extra_sections return false.

ld/
	PR ld/34448
	* ldlang.c (lang_gc_sections): Check bfd_gc_sections return and
	report the fatal error.

Signed-off-by: H.J. Lu <hjl.tools@gmail.com>
Don't dump GOT section contents after seeing errors in input:

readelf: Error: Section 10 has invalid sh_entsize of 0
readelf: Error: (Using the expected size of 18 for the rest of this dump)
readelf: Error: Too many program headers - 0x3030 - the file is not that big

	PR binutils/34473
	* elfcomm.c (seen_error): New.
	(seen_elf_error): Likewise.
	(clear_elf_error): Likewise.
	(error): Set seen_error to true.
	* elfcomm.h (seen_elf_error): New.
	(clear_elf_error): Likewise.
	* readelf.c (process_got_section_contents): Return false if
	seen_elf_error returns true.
	(main): Call clear_elf_error before calling process_file.

Signed-off-by: H.J. Lu <hjl.tools@gmail.com>
Now that the GDB 18 branch has been created,
this commit bumps the version number in gdb/version.in to
19.0.50.DATE-git

For the record, the GDB 18 branch was created
from commit b737567.
This commit a new section for the next release branch, and renames
the section of the current branch, now that it has been cut.
check_types_equal compares both is_nottext and instance_flags, but the
latter includes the former, so the redundant check can be removed.

Reviewed-By: Tankut Baris Aktemur <TankutBaris.Aktemur@amd.com>
Approved-By: Tom de Vries <tdevries@suse.de>
erratum_84319_opts starts at ERRAT_NONE = (1 << 0), so a plain boolean test
on fix_erratum_843419 is true even when no erratum workaround was asked for.
Every other use in the file tests against ERRAT_NONE or masks with ERRAT_ADR /
ERRAT_ADRP.  Two do not.

The bare test dates from the conversion of fix_erratum_843419 from an int to an
enum for PR ld/24373.

Having the workaround on by default all the time is, of course, undesirable as
it costs link-time and is not what the user has asked by default.

Tested on aarch64-none-linux-gnu.

bfd/
	* elfnn-aarch64.c (elfNN_aarch64_write_section): Test
	fix_erratum_843419 against ERRAT_NONE.
	(elfNN_aarch64_late_size_sections): Likewise.

Signed-off-by: Kyrylo Tkachov <ktkachov@nvidia.com>
@amd-shahab amd-shahab self-assigned this Aug 7, 2026
@amd-shahab
amd-shahab requested a review from a team as a code owner August 7, 2026 16:15
@aktemur

aktemur commented Aug 7, 2026

Copy link
Copy Markdown
Contributor

The merge looks clean. Do we want to add XFAILs as part of this PR? That would make sense to me to avoid intermediary regressions.

@lumachad

lumachad commented Aug 7, 2026

Copy link
Copy Markdown
Collaborator

If these failures are expected from rocgdb's testsuite's side, we need to make them XFAIL rather than force CI to ignore them.

Which is it?

@lancesix

lancesix commented Aug 7, 2026

Copy link
Copy Markdown
Collaborator

The merge looks clean. Do we want to add XFAILs as part of this PR? That would make sense to me to avoid intermediary regressions.

gfx1103 is "new", so not really a regression, but would be good to add the XFAIL in watchpoint-basic before pulling from upstream (https://github.com/ROCm/ROCgdb/blob/amd-staging/gdb/testsuite/gdb.rocm/watchpoint-basic.exp#L99)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

8 participants