[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Paul Floyd]

https://bugs.kde.org/show_bug.cgi?id=365327

Paul Floyd  changed:

   What|Removed |Added

 CC||pjfl...@wanadoo.fr

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

Rhys Kidd  changed:

   What|Removed |Added

 Depends on||379893


Referenced Bugs:

https://bugs.kde.org/show_bug.cgi?id=379893
[Bug 379893] valgrind segfault os x 10.12.4 (_pthread_body)
-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327
Bug 365327 depends on bug 379754, which changed state.

Bug 379754 Summary: [PATCH] Missing syscall ulock_wait (OS X 10.12)
https://bugs.kde.org/show_bug.cgi?id=379754

   What|Removed |Added

 Status|CONFIRMED   |RESOLVED
 Resolution|--- |FIXED

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327
Bug 365327 depends on bug 385279, which changed state.

Bug 385279 Summary: [PATCH] unhandled syscall: mach:43 
(mach_generate_activity_id)
https://bugs.kde.org/show_bug.cgi?id=385279

   What|Removed |Added

 Status|CONFIRMED   |RESOLVED
 Resolution|--- |FIXED

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

Rhys Kidd  changed:

   What|Removed |Added

 Depends on|383723  |


Referenced Bugs:

https://bugs.kde.org/show_bug.cgi?id=383723
[Bug 383723] SIGILL failure with ud2 opcode (macOS)
-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

Rhys Kidd  changed:

   What|Removed |Added

 Depends on||385279


Referenced Bugs:

https://bugs.kde.org/show_bug.cgi?id=385279
[Bug 385279] unhandled syscall: mach:43 (mach_generate_activity_id)
-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327
Bug 365327 depends on bug 379373, which changed state.

Bug 379373 Summary: [PATCH] Syscall param msg->desc.port.name points to 
uninitialised byte(s) on macOS 10.12
https://bugs.kde.org/show_bug.cgi?id=379373

   What|Removed |Added

 Status|CONFIRMED   |RESOLVED
 Resolution|--- |FIXED

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327
Bug 365327 depends on bug 379748, which changed state.

Bug 379748 Summary: [PATCH] Missing pselect syscall (OS X 10.11)
https://bugs.kde.org/show_bug.cgi?id=379748

   What|Removed |Added

 Status|CONFIRMED   |RESOLVED
 Resolution|--- |FIXED

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

Rhys Kidd  changed:

   What|Removed |Added

 Depends on||383723


Referenced Bugs:

https://bugs.kde.org/show_bug.cgi?id=383723
[Bug 383723] MacOS 10.12.x: UNKNOWN workq_ops option 128, and ud2 opcode
-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Louis Brunner]

https://bugs.kde.org/show_bug.cgi?id=365327

Louis Brunner  changed:

   What|Removed |Added

 Depends on||380269


Referenced Bugs:

https://bugs.kde.org/show_bug.cgi?id=380269
[Bug 380269] No multithreading in macOS Sierra (10.12)
-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Louis Brunner]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #26 from Louis Brunner  ---
Created attachment 105483
  --> https://bugs.kde.org/attachment.cgi?id=105483=edit
Checks in new syscalls/traps

Rhys,

I created new issue for pselect and ulock_wait (couldn't find the failing test
though).
Could we simply disable pth_term_signal (the pthread test that hang) while we
search for a fix? It makes testing pretty cumbersome. Do you have any idea what
the problem could be? Maybe linked to another pthread problem (like 349128)?

I have also included a patch based on the one for ulock_wait (379754) which
includes more checks for the newly added syscalls/traps. Do you want me to
create an issue per syscall?

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Louis Brunner]

https://bugs.kde.org/show_bug.cgi?id=365327

Louis Brunner  changed:

   What|Removed |Added

 Depends on||379754


Referenced Bugs:

https://bugs.kde.org/show_bug.cgi?id=379754
[Bug 379754] Missing syscall ulock_wait (OS X 10.12)
-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Louis Brunner]

https://bugs.kde.org/show_bug.cgi?id=365327

Louis Brunner  changed:

   What|Removed |Added

 Depends on||379748


Referenced Bugs:

https://bugs.kde.org/show_bug.cgi?id=379748
[Bug 379748] Missing pselect syscall
-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #25 from Rhys Kidd  ---
I'd strongly suggest running through the regression test suite on your local
setup to provide coverage of a reasonable amount of the Darwin code base. It is
run with 'make regtest'.

The regression test is not quite automated on Darwin. You will probably need to
kill one of the pthread-related tests that hangs at present.

Suggested areas to focus on:
1. pselect syscall
2. Any other missing syscalls that prevent you using a program of interest to
you - more motivated to fix bugs in open source you're personally motivated in!
3. ulock_wait syscall - grep through the regression tests *.stderr.diff for the
test that triggers it
4. Find a fix for the pthread-related hang in the regression test suite.

Create a separate bug report for each, then attach a patch and make it a
blocker for this bug report.

The compiler warnings for 'getentropy' and 'ulock_wake' were fixed before
commiting to SVN.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Louis Brunner]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #24 from Louis Brunner  ---
Hi Rhys,

Thank you for the merge!

While making the changes for the new dylib loading, I encountered a crash
coming from is_in_syscall.
I blamed it on the experimental nature of my patch at the time and I can't
remember what triggered the crash.

You mentioned that you had older macOS VMs, do you have any advice/guide of any
sort to help me set one up for testing?
I could look into the failing test and the warnings linked to `getentropy` and
`ulock_wake` (if you didn't fix them already).

I was also working on a new version that included a pselect syscall wrapper
(found it missing while testing python with valgrind), I will add it later.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #23 from Rhys Kidd  ---
So some progress has been made. As of r16328 Valgrind SVN trunk on macOS 10.12
is reporting as follows via its regression test suite:

== 629 tests, 450 stderr failures, 79 stdout failures, 1 stderrB failure, 1
stdoutB failure, 31 post failures ==

Whilst this might seem poor, it's only about 200 stderr failures above a
similar OS X 10.11 system. And the regressions are predominantly caused solely
by bz#379373.

Getting that fixed should greatly improve the support of Valgrind on this
modern macOS. Thanks for the patches!

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327
Bug 365327 depends on bug 379372, which changed state.

Bug 379372 Summary: UNKNOWN task message [id 3447, to mach_task_self(), reply 
0x603] (task_register_dyld_shared_cache_image_info)
https://bugs.kde.org/show_bug.cgi?id=379372

   What|Removed |Added

 Status|CONFIRMED   |RESOLVED
 Resolution|--- |FIXED

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327
Bug 365327 depends on bug 379371, which changed state.

Bug 379371 Summary: UNKNOWN task message [id 3444, to mach_task_self(), reply 
0x603] (task_register_dyld_image_infos)
https://bugs.kde.org/show_bug.cgi?id=379371

   What|Removed |Added

 Status|CONFIRMED   |RESOLVED
 Resolution|--- |FIXED

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327
Bug 365327 depends on bug 379390, which changed state.

Bug 379390 Summary: unhandled syscall: mach:70 (host_create_mach_voucher_trap)
https://bugs.kde.org/show_bug.cgi?id=379390

   What|Removed |Added

 Status|CONFIRMED   |RESOLVED
 Resolution|--- |FIXED

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #22 from Rhys Kidd  ---
Regarding outstanding patch #2 above in my Comment 20, I've now been able to
reproduce the missing mach syscall 70 within Valgrind's regression test suite.

bz#379390 is tracking that issue.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

Rhys Kidd  changed:

   What|Removed |Added

 Depends on||379390


Referenced Bugs:

https://bugs.kde.org/show_bug.cgi?id=379390
[Bug 379390] unhandled syscall: mach:70 (host_create_mach_voucher_trap)
-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

Rhys Kidd  changed:

   What|Removed |Added

 Attachment #105218|0   |1
is obsolete||

--- Comment #21 from Rhys Kidd  ---
Created attachment 105282
  --> https://bugs.kde.org/attachment.cgi?id=105282=edit
0001-is_in_syscall-assert.patch

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #20 from Rhys Kidd  ---
Louis Brunner,

Thank you for the macOS 10.12 patch to improve Valgrind's support on that
platform. You may have seen that I've committed most of the changes to the SVN
master version today.

A short summary follows, including the items that haven't been committed and
why:

Committed to SVN master
1. Improve consistency of VG_(printf)() usage in coregrind/m_ume/macho.c
(r16317)
2. New macOS 10.12 way of loading dylib (r16318 - biggest improvement)
3. Add a no-op wrapper for a new-in-10.12 syscall: getentropy (r16319)
4. Add a no-op wrapper for a new-in-10.12 syscall: ulock_wake (r16320)
5. Add a no-op wrapper for a new-in-10.10 syscall: csrctl (r16321)
6. Add a no-op wrapper for a new-in-10.10 syscall: faccessat (r16322)
7. Add a no-op wrapper for a new-in-10.10 syscall: fstatat64 (r16323)

Patches not committed and to discuss:

1. is_in_syscall-assert as noted above this caused one regression. The rebased,
standalone patch attached is to be investigated further.

2. host_create_mach_voucher_trap or VG_DARWIN_SYSCALL_CONSTRUCT_MACH(70). Can
you provide details of the reproducible bug this is fixing? I was unable to
identify a fix this caused. Note we already have a host_create_mach_voucher()
implemented.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

Rhys Kidd  changed:

   What|Removed |Added

 Depends on||379373


Referenced Bugs:

https://bugs.kde.org/show_bug.cgi?id=379373
[Bug 379373] Syscall param msg->desc.port.name points to uninitialised byte(s)
on macOS 10.12
-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

Rhys Kidd  changed:

   What|Removed |Added

 Depends on||379372


Referenced Bugs:

https://bugs.kde.org/show_bug.cgi?id=379372
[Bug 379372] UNKNOWN task message [id 3447, to mach_task_self(), reply 0x603]
(task_register_dyld_shared_cache_image_info)
-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

Rhys Kidd  changed:

   What|Removed |Added

 Depends on||379371


Referenced Bugs:

https://bugs.kde.org/show_bug.cgi?id=379371
[Bug 379371] UNKNOWN task message [id 3444, to mach_task_self(), reply 0x603]
(task_register_dyld_image_infos)
-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #19 from Rhys Kidd  ---
There are some new warnings due to scope issues with the patch as is, i.e. see
the below warnings that creep in on pre macOS 10.12 systems [0].

> m_syswrap/syswrap-darwin.c:9781:1: warning: no previous prototype for 
> function 'vgSysWrap_darwin_getentropy_before' [-Wmissing-prototypes]
> PRE(getentropy)
> ^
> m_syswrap/syswrap-darwin.c:9788:1: warning: no previous prototype for 
> function 'vgSysWrap_darwin_ulock_wake_before' [-Wmissing-prototypes]
> PRE(ulock_wake)
> ^

Before I commit this, the patch v2 will be broken up into smaller discrete
patches to ease bisecting later on.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #18 from Rhys Kidd  ---
With this patch I saw a regression in massif/tests/mmapunmap on macOS 10.11,
due to the vg_assert(syscallInfo) added in VG_(is_in_syscall)().

For now I'm going to hold off committing that specific part of your v2 patch,
although you could well have uncovered an existing bug that is only being
tickled with this change. 

--- mmapunmap.stderr.exp2016-04-05 21:07:04.0 -0400
+++ mmapunmap.stderr.out2017-04-29 14:42:38.0 -0400
@@ -0,0 +1,19 @@
+
+valgrind: m_syswrap/syswrap-main.c:1691 (Bool vgPlain_is_in_syscall(Int)):
Assertion 'syscallInfo' failed.
+
+host stacktrace:
+   at 0x238014AA3: ???
+   by 0x238014E4C: ???
+   by 0x238014E2A: ???
+   by 0x2380B1228: ???
+   by 0x238030C77: ???
+   by 0x238031030: ???
+   by 0x238003A7A: ???
+   by 0x2380038AE: ???
+   by 0x23801DADC: ???
+   by 0x23801C943: ???
+
+sched status:
+  running_tid=1
+
+Thread 1: status = VgTs_Init (lwpid 0)

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #17 from Rhys Kidd  ---
I'll test this patch on my 10.10 and 10.9 VMs.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Louis Brunner]

https://bugs.kde.org/show_bug.cgi?id=365327

Louis Brunner  changed:

   What|Removed |Added

 Attachment #104964|0   |1
is obsolete||

--- Comment #16 from Louis Brunner  ---
Created attachment 105218
  --> https://bugs.kde.org/attachment.cgi?id=105218=edit
macOS Sierra incomplete support v2

Thank you for the feedback!

Austin, I just added a new patch that solves this issue (I can't believe that I
didn't check that...). Thanks for the check on 10.11!

Julian, I don't have access to any 10.10 or 10.11, but I am trying to install
one in a VM right now. Technically, most of the changes are scoped in the
proper `#if DARWIN_VERS >= DARWIN_10_XX`.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Austin English]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #15 from Austin English  ---
(In reply to Louis Brunner from comment #12)
> Created attachment 104964 [details]
> macOS Sierra incomplete support
> 
> I have been working on a patch to support macOS Sierra.
> At the moment, it works for a variety of programs (python, GIMP, most basic
> commands like ls, mkdir) but it is still incomplete (warnings in dyld,
> libsystem_kernel... crash for some GUI programs).
> 
> The patch adds a few required syscalls with placeholder implementations
> (faccessat, fstatat64, csrctl, getentropy and ulock_wake) and the new way of
> loading dylib (placing them at the end of the currently loaded segments).
> The second change means we need to know where the last segment was loaded,
> which means carrying around one more pointer on pointer in every function
> (which already have 6-9 arguments), that's why I created a structure
> (load_info_t) to store all this information and easily carry it around.
> 
> It also adds one assert in is_in_syscall in
> coregrind/m_syswrap/syswrap-main.c to match the other syscall related
> functions in the same file. I had a difficult to diagnose crash in this
> function because it didn't check for the existence of the syscall table.
> 
> Tell me if you need any change

I can confirm that the situation is improved on 10.12 with this patch. E.g.,
valgrind will at least attempt things now. I noticed that `./vg-in-place ls
-al` crashes valgrind, though `./vg-in-place true` works fine (does show some
uninitialized memory in OSX though).

A basic sanity check on 10.11 also worked for me.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Julian Seward]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #14 from Julian Seward  ---
(In reply to Louis Brunner from comment #12)

Louis, thanks for working on this.  A question: with your patch applied,
does it still work on earlier MacOS versions (10.11, 10.10) ?

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #13 from julien.aubert.m...@gmail.com ---
Thanks. Works fine for me so far!

2017-04-11 12:40 GMT+02:00 Louis Brunner :

> https://bugs.kde.org/show_bug.cgi?id=365327
>
> Louis Brunner  changed:
>
>What|Removed |Added
> 
> 
>  CC||
> louis.brunner...@gmail.com
>
> --- Comment #12 from Louis Brunner  ---
> Created attachment 104964
>   --> https://bugs.kde.org/attachment.cgi?id=104964=edit
> macOS Sierra incomplete support
>
> I have been working on a patch to support macOS Sierra.
> At the moment, it works for a variety of programs (python, GIMP, most basic
> commands like ls, mkdir) but it is still incomplete (warnings in dyld,
> libsystem_kernel... crash for some GUI programs).
>
> The patch adds a few required syscalls with placeholder implementations
> (faccessat, fstatat64, csrctl, getentropy and ulock_wake) and the new way
> of
> loading dylib (placing them at the end of the currently loaded segments).
> The
> second change means we need to know where the last segment was loaded,
> which
> means carrying around one more pointer on pointer in every function (which
> already have 6-9 arguments), that's why I created a structure
> (load_info_t) to
> store all this information and easily carry it around.
>
> It also adds one assert in is_in_syscall in coregrind/m_syswrap/syswrap-
> main.c
> to match the other syscall related functions in the same file. I had a
> difficult to diagnose crash in this function because it didn't check for
> the
> existence of the syscall table.
>
> Tell me if you need any change
>
> --
> You are receiving this mail because:
> You voted for the bug.
>

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Louis Brunner]

https://bugs.kde.org/show_bug.cgi?id=365327

Louis Brunner  changed:

   What|Removed |Added

 CC||louis.brunner...@gmail.com

--- Comment #12 from Louis Brunner  ---
Created attachment 104964
  --> https://bugs.kde.org/attachment.cgi?id=104964=edit
macOS Sierra incomplete support

I have been working on a patch to support macOS Sierra.
At the moment, it works for a variety of programs (python, GIMP, most basic
commands like ls, mkdir) but it is still incomplete (warnings in dyld,
libsystem_kernel... crash for some GUI programs).

The patch adds a few required syscalls with placeholder implementations
(faccessat, fstatat64, csrctl, getentropy and ulock_wake) and the new way of
loading dylib (placing them at the end of the currently loaded segments). The
second change means we need to know where the last segment was loaded, which
means carrying around one more pointer on pointer in every function (which
already have 6-9 arguments), that's why I created a structure (load_info_t) to
store all this information and easily carry it around.

It also adds one assert in is_in_syscall in coregrind/m_syswrap/syswrap-main.c
to match the other syscall related functions in the same file. I had a
difficult to diagnose crash in this function because it didn't check for the
existence of the syscall table.

Tell me if you need any change

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[chrstphrchvz]

https://bugs.kde.org/show_bug.cgi?id=365327

chrstphrchvz  changed:

   What|Removed |Added

 CC||chrischa...@gmx.us

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[vladimir]

https://bugs.kde.org/show_bug.cgi?id=365327

vladimir  changed:

   What|Removed |Added

 CC||vla...@scatt.com

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[René Schwaiger]

https://bugs.kde.org/show_bug.cgi?id=365327

René Schwaiger  changed:

   What|Removed |Added

 CC||sansseco...@me.com

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Austin English]

https://bugs.kde.org/show_bug.cgi?id=365327

Austin English  changed:

   What|Removed |Added

 CC||austinengl...@gmail.com

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Anton Korobeynikov]

https://bugs.kde.org/show_bug.cgi?id=365327

Anton Korobeynikov  changed:

   What|Removed |Added

 CC||an...@korobeynikov.info

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[zwing99]

https://bugs.kde.org/show_bug.cgi?id=365327

zwing99  changed:

   What|Removed |Added

 CC||zac.o...@gmail.com

--- Comment #11 from zwing99  ---
Re: Bartel's comment is there any progress on this?

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Bartel]

https://bugs.kde.org/show_bug.cgi?id=365327

Bartel  changed:

   What|Removed |Added

 CC||bartel.eerdek...@luminex.be

--- Comment #10 from Bartel  ---
Any idea when Sierra support can be expected?

Thanks!

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[CInk_Ibrahim]

https://bugs.kde.org/show_bug.cgi?id=365327

CInk_Ibrahim  changed:

   What|Removed |Added

 CC||ibra...@cinkstone.com

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[hmijail]

https://bugs.kde.org/show_bug.cgi?id=365327

hmijail  changed:

   What|Removed |Added

 CC||hmij...@gmail.com

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[MrCraftCod]

https://bugs.kde.org/show_bug.cgi?id=365327

MrCraftCod  changed:

   What|Removed |Added

 CC||zerd...@gmail.com

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Mark Abraham]

https://bugs.kde.org/show_bug.cgi?id=365327

Mark Abraham  changed:

   What|Removed |Added

 CC||mark.j.abra...@gmail.com

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=365327

plabo...@gmail.com changed:

   What|Removed |Added

 CC||plabo...@gmail.com

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #9 from Rhys Kidd  ---
Apple have just this week finally released the source code for critical system
components that shipped in macOS Sierra (10.12). This includes xnu, the kernel.

https://opensource.apple.com/release/macos-1012.html

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

Rhys Kidd  changed:

   What|Removed |Added

 CC||zalbi...@gmail.com

--- Comment #8 from Rhys Kidd  ---
*** Bug 372772 has been marked as a duplicate of this bug. ***

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Bjørn D . Rasmussen]

https://bugs.kde.org/show_bug.cgi?id=365327

Bjørn D. Rasmussen  changed:

   What|Removed |Added

 CC||bjoern.d.rasmus...@gmail.co
   ||m

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #7 from Rhys Kidd  ---
Thanks for testing the code and reporting here.

Yes, the preliminary support is just that. There's known issues, including this
one that appears to relate to the way NULL pages are mapped.

Unfortunately, Apple is still yet to release their APSL-licensed kernel xnu
from Sierra as source code (have written their open source team two emails thus
far with no response). It makes solving it harder than it should be.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[joaodlf]

https://bugs.kde.org/show_bug.cgi?id=365327

joaodlf  changed:

   What|Removed |Added

 CC||jdlferreir...@gmail.com

--- Comment #6 from joaodlf  ---
I'm running macOS (10.12.1) and valgrind-devel @3.13.0-r16104 (devel) and
facing the following error when running valgrind --leak-check=yes on a very
simple executable:

valgrind: mmap-FIXED(0x0, 253952) failed in UME (load_segment1) with error 12
(Cannot allocate memory).

I'm going to assume "preliminary support" doesn't quite get us there yet :).

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Julian Seward via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #5 from Julian Seward  ---
(In reply to Rhys Kidd from comment #4)
> Preliminary support added in r15976.

Merged to 3_12_BRANCH in r16071.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #4 from Rhys Kidd  ---
Preliminary support added in r15976.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=365327

Rhys Kidd  changed:

   What|Removed |Added

 Attachment #101036|0   |1
is obsolete||

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Julian Seward via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #3 from Julian Seward  ---
The patches look OK to me.  Only one nit:

--- coregrind/m_syswrap/syswrap-amd64-darwin.c(revision 15949)
+++ coregrind/m_syswrap/syswrap-amd64-darwin.c(working copy)
+#  elif DARWIN_VERS == DARWIN_10_9 || DARWIN_VERS == DARWIN_10_10 ||
DARWIN_VERS == DARWIN_10_11 || DARWIN_VERS == DARWIN_10_12

Please break such long lines so as to remain within 80 cols where possible.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=365327

Rhys Kidd  changed:

   What|Removed |Added

 Ever confirmed|0   |1
 Status|UNCONFIRMED |CONFIRMED

--- Comment #2 from Rhys Kidd  ---
Initial patch set attached that brings up macOS 10.12 support.

Patches against r15949, that allow valgrind and test suite to build on macOS
10.12.

Valgrind binaries do not yet run on macOS 10.12, due to an error mapping the
NULL page during load of /usr/lib/dyld, see output below:

$ uname -v
Darwin Kernel Version 16.0.0: Tue Aug 23 17:02:44 PDT 2016;
root:xnu-3789.1.31~2/RELEASE_X86_64
$ ./vg-in-place -d -d ls
...
--42669:1: initimg Loading client
--42669:1: ume load_thin_file: begin:   /bin/ls
--42669:2: ume   mmap fixed (file) (0x1, 20480)
--42669:2: ume   mmap fixed (file) (0x15000, 4096)
--42669:2: ume   mmap fixed (file) (0x16000, 16384)
--42669:1: ume load_thin_file: begin:   /usr/lib/dyld
--42669:2: ume   mmap fixed (file) (0x0, 253952)
valgrind: mmap-FIXED(0x0, 253952) failed in UME (load_segment1) with error 12
(Cannot allocate memory).

On OS X 10.11, that equivalent last mmap is at a higher base address:

$ uname -v
Darwin Kernel Version 15.6.0: Thu Jun 23 18:25:34 PDT 2016;
root:xnu-3248.60.10~1/RELEASE_X86_64
$ ./vg-in-place -d -d ls
...
--64933:1: initimg Loading client
--64933:1: ume load_thin_file: begin:   /bin/ls
--64933:2: ume   mmap fixed (file) (0x1, 20480)
--64933:2: ume   mmap fixed (file) (0x15000, 4096)
--64933:2: ume   mmap fixed (file) (0x16000, 16384)
--64933:1: ume load_thin_file: begin:   /usr/lib/dyld
--64933:2: ume   mmap fixed (file) (0x7fff5fc0, 229376)
...

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=365327

--- Comment #1 from Rhys Kidd  ---
Created attachment 101036
  --> https://bugs.kde.org/attachment.cgi?id=101036=edit
Initial support for macOS 10.12 (incomplete) patch set

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=365327
Bug 365327 depends on bug 366138, which changed state.

Bug 366138 Summary: configure errors out when using Xcode 8 (clang 8.0.0)
https://bugs.kde.org/show_bug.cgi?id=366138

   What|Removed |Added

 Status|CONFIRMED   |RESOLVED
 Resolution|--- |FIXED

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=365327

Rhys Kidd  changed:

   What|Removed |Added

 Depends on||366138

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 365327] Support macOS Sierra (10.12)

[Rhys Kidd via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=365327

Rhys Kidd  changed:

   What|Removed |Added

   Assignee|jsew...@acm.org |rhysk...@gmail.com

-- 
You are receiving this mail because:
You are watching all bug changes.