Re: 2.39 Release Schedule

On Mon, Jun 27, 2022 at 08:43:46PM +0200, Matthias Klose wrote:
> On 27.06.22 13:37, Nick Clifton via Binutils wrote:
> > Hi Guys,
> > 
> >    OK, so the plan is that the branch for the 2.39 release will be
> >    cut on Friday July 8th (just before I go on vacation).  All being
> >    well the release itself will happen in the first weekend of August
> >    (Sat 6th probably).
> > 
> >    If you have new features that you want to get into the 2.39 release
> >    now is the time to submit them - or ping for a review if they have
> >    already been submitted.
> > 
> >    Once the branch is cut the usual rules will apply - bug fixes and
> >    doc updates only - although this time around I am also going to ask
> >    that any maintainer who wishes to do so review and approve/reject
> >    patches for the branch.
> > 
> >    I will be away July 11..July 22 but as soon as I get back I will be
> >    on patch review duty for as much of the time as I can.
> 
> with today's trunk, I see regressions on i686-linux-gnu and mips64el-linux-gnu:
> 
> i386:
> W: [ld-elf/elf.exp] REGRESSION (UNSUPPORTED -> FAIL): static init array mixed
> 
> trunk 20220622 didn't show that.
> 
> mips64el:
> I: [binutils-all/objcopy.exp] progression (UNTESTED -> PASS): simple objcopy
> of executable
> I: [binutils-all/objcopy.exp] progression (UNTESTED -> PASS): run objcopy of
> executable
> I: [binutils-all/objcopy.exp] progression (UNTESTED -> PASS): run stripped
> executable
> I: [binutils-all/objcopy.exp] progression (UNTESTED -> PASS): run stripped
> executable with saving a symbol
> I: [binutils-all/objcopy.exp] progression (UNTESTED -> PASS): keep only debug data
> I: [binutils-all/objcopy.exp] progression (UNTESTED -> PASS): simple objcopy
> of debug data
> I: [binutils-all/objcopy.exp] progression (UNTESTED -> PASS): NOBITS
> sections retain sh_link field

All of the above might be explained by changes in your testing
environment.  eg. Installing compilers and libraries for the affected
targets.

> W: [ld-elf/elf.exp] REGRESSION (PASS -> FAIL): PR ld/29072 (warn about an
> executable .note.GNU-stack)

	PR 29263
	* configure.ac: Fix typo.
	* testsuite/ld-elf/elf.exp: Add mips to targets that need
	--warn-execstack to pass first pr29072 test.

diff --git a/ld/configure.ac b/ld/configure.ac
index a54a2801889..4331d6b1302 100644
--- a/ld/configure.ac
+++ b/ld/configure.ac
@@ -564,7 +564,7 @@ AC_DEFINE_UNQUOTED(DEFAULT_LD_WARN_EXECSTACK,
   [Define to 1 if you want to enable --warn-execstack in ELF linker by default.])
 
 if test "${ac_default_ld_warn_rwx_segments}" = unset; then
-  ac_default_ld_warn_rwx_segments=1 ;;
+  ac_default_ld_warn_rwx_segments=1
 fi
 AC_DEFINE_UNQUOTED(DEFAULT_LD_WARN_RWX_SEGMENTS,
   $ac_default_ld_warn_rwx_segments,
diff --git a/ld/testsuite/ld-elf/elf.exp b/ld/testsuite/ld-elf/elf.exp
index 473349dd0b2..10c635e9593 100644
--- a/ld/testsuite/ld-elf/elf.exp
+++ b/ld/testsuite/ld-elf/elf.exp
@@ -238,8 +238,8 @@ if {   [istarget *-*-*linux*]
 
     # The HPPA target disables the warnings by default since it needs an
     # executable stack to support older kernels and an RWX segment to
-    # support its PLT.
-    if { [istarget "hppa*-*-*"] } {
+    # support its PLT.  MIPS also turns off exec stack warnings.
+    if { [istarget "hppa*-*-*"] || [istarget mips*-*-*] } {
         set extra_link_opts "--warn-execstack --warn-rwx-segments"
     } else {
         set extra_link_opts ""


-- 
Alan Modra
Australia Development Lab, IBM