emaste (Ed Maste)
User

Projects

User Details

User Since
Jul 24 2013, 5:36 AM (238 w, 6 d)

Recent Activity

Yesterday

emaste added a comment to D43168: Don't use -ldl on BSD.

In FreeBSD libpthread.so is a symlink to libthr; -pthread and -lpthread should both work.

Mon, Feb 19, 9:23 PM

Fri, Feb 16

emaste added a comment to D43370: Add initial XRay support for NetBSD.

This LGTM

Fri, Feb 16, 6:46 AM · Restricted Project

Fri, Feb 9

emaste committed rL324739: Add IMPLEMENTATION NOTES describing lld's .a handling in the man page.
Add IMPLEMENTATION NOTES describing lld's .a handling in the man page
Fri, Feb 9, 7:38 AM
emaste committed rLLD324739: Add IMPLEMENTATION NOTES describing lld's .a handling in the man page.
Add IMPLEMENTATION NOTES describing lld's .a handling in the man page
Fri, Feb 9, 7:38 AM
emaste closed D43072: ld.lld.1: add IMPLEMENTATION NOTES section documenting lld's handling.
Fri, Feb 9, 7:38 AM

Thu, Feb 8

emaste updated subscribers of D43072: ld.lld.1: add IMPLEMENTATION NOTES section documenting lld's handling.
Thu, Feb 8, 5:47 PM
emaste created D43072: ld.lld.1: add IMPLEMENTATION NOTES section documenting lld's handling.
Thu, Feb 8, 7:16 AM

Wed, Feb 7

emaste committed rL324528: Document --build-id=fast in ld.lld.1.
Document --build-id=fast in ld.lld.1
Wed, Feb 7, 1:24 PM
emaste committed rLLD324528: Document --build-id=fast in ld.lld.1.
Document --build-id=fast in ld.lld.1
Wed, Feb 7, 1:24 PM
emaste closed D43039: ld.lld.1: document --build-id=fast.
Wed, Feb 7, 1:24 PM
emaste created D43039: ld.lld.1: document --build-id=fast.
Wed, Feb 7, 1:05 PM
emaste added inline comments to D42963: Add initial ld.lld.1 man page.
Wed, Feb 7, 12:45 PM
emaste committed rL324512: Add ld.lld.1 man page, describing lld's ELF linker.
Add ld.lld.1 man page, describing lld's ELF linker
Wed, Feb 7, 12:40 PM
emaste committed rLLD324512: Add ld.lld.1 man page, describing lld's ELF linker.
Add ld.lld.1 man page, describing lld's ELF linker
Wed, Feb 7, 12:40 PM
emaste closed D42963: Add initial ld.lld.1 man page.
Wed, Feb 7, 12:40 PM
emaste added a comment to D42963: Add initial ld.lld.1 man page.

Thanks, I'll commit this later today. Not sure how I ended up with 01 instead of O1 but have fixed those locally.

Wed, Feb 7, 12:08 PM
emaste added inline comments to D42963: Add initial ld.lld.1 man page.
Wed, Feb 7, 6:21 AM
emaste added inline comments to D42963: Add initial ld.lld.1 man page.
Wed, Feb 7, 6:16 AM
emaste added a comment to D42963: Add initial ld.lld.1 man page.

PDF version of current version:

Wed, Feb 7, 6:16 AM
emaste updated the diff for D42963: Add initial ld.lld.1 man page.
Wed, Feb 7, 6:16 AM

Tue, Feb 6

emaste added inline comments to D42963: Add initial ld.lld.1 man page.
Tue, Feb 6, 6:04 PM
emaste added a comment to D42963: Add initial ld.lld.1 man page.

PDF rendered version of the man page:

Tue, Feb 6, 6:04 PM
emaste added inline comments to D42963: Add initial ld.lld.1 man page.
Tue, Feb 6, 6:01 PM
emaste updated the diff for D42963: Add initial ld.lld.1 man page.

Incorporate review feedback

Tue, Feb 6, 5:59 PM
emaste updated the diff for D42963: Add initial ld.lld.1 man page.
  • Update copyright and license statement
  • Remove internal -full-shutdown option
Tue, Feb 6, 9:40 AM
emaste created D42963: Add initial ld.lld.1 man page.
Tue, Feb 6, 6:51 AM

Fri, Jan 26

emaste added inline comments to D42550: Put the header in the first PT_LOAD even if that PT_LOAD has a LMAExpr.
Fri, Jan 26, 6:35 AM

Thu, Jan 25

emaste added a comment to D31528: [ELF][MIPS] Multi-GOT implementation.

We have a chance to celebrate the first anniversary of this patch in a couple of months. Is it possible to escape this dubious holiday?

Thu, Jan 25, 2:10 PM · lld
emaste added inline comments to D42550: Put the header in the first PT_LOAD even if that PT_LOAD has a LMAExpr.
Thu, Jan 25, 11:20 AM

Jan 11 2018

emaste accepted D41635: Align SHT_NOBITS sections is they are the first on a PT_LOAD.

Seems fine to me

Jan 11 2018, 7:37 PM

Jan 10 2018

emaste added a comment to D41930: [Fuchsia] Use llvm-objcopy as objcopy on non-Apple hosts.

The title is more correctly "non-Apple hosts"? I.e., building on FreeBSD will also use llvm-objcopy.

Jan 10 2018, 5:37 PM
emaste committed rLLD322180: Correct typo in help text.
Correct typo in help text
Jan 10 2018, 4:56 AM
emaste committed rL322180: Correct typo in help text.
Correct typo in help text
Jan 10 2018, 4:56 AM

Jan 4 2018

emaste added a comment to D41723: Introduce the "retpoline" x86 mitigation technique for variant #2 of the speculative execution vulnerabilities disclosed today, specifically identified by CVE-2017-5715, "Branch Target Injection", and is one of the two halves to Spectre...

Is there any plan (or willingness?) to backport this further, to 4.0.0 and even 3.4.1?

Jan 4 2018, 11:40 AM
emaste updated subscribers of D41723: Introduce the "retpoline" x86 mitigation technique for variant #2 of the speculative execution vulnerabilities disclosed today, specifically identified by CVE-2017-5715, "Branch Target Injection", and is one of the two halves to Spectre...
Jan 4 2018, 4:42 AM

Dec 29 2017

emaste added a comment to D41592: [ELF] - Add missing dynamic tags when producing output with IRelative relocations only..

With this patch applied Kostik's original testcase passes.

Dec 29 2017, 1:27 PM
emaste added inline comments to D41592: [ELF] - Add missing dynamic tags when producing output with IRelative relocations only..
Dec 29 2017, 10:13 AM

Dec 26 2017

emaste accepted D41551: [ELF] - Allow relocation to a weak undefined symbol when -z notext is given..

LGTM

Dec 26 2017, 1:33 PM
emaste added inline comments to D41551: [ELF] - Allow relocation to a weak undefined symbol when -z notext is given..
Dec 26 2017, 6:27 AM
emaste added a comment to D41551: [ELF] - Allow relocation to a weak undefined symbol when -z notext is given..

The original FreeBSD test case (a 'hello world' linked with lld -znotext) works with this change applied.

Dec 26 2017, 6:26 AM

Dec 25 2017

emaste added a comment to D41551: [ELF] - Allow relocation to a weak undefined symbol when -z notext is given..

Conflicts after rL321430. I tried a straightforward rebase (moving the if ((S.Flags & SHF_WRITE) || !Config->ZText) block below if (Sym.isUndefWeak()) but test/ELF/retain-und.s (from rL293093) fails.

Dec 25 2017, 7:31 PM

Dec 24 2017

emaste added a comment to D41551: [ELF] - Allow relocation to a weak undefined symbol when -z notext is given..

LGTM.

Dec 24 2017, 10:01 AM

Dec 23 2017

emaste added a comment to D41551: [ELF] - Allow relocation to a weak undefined symbol when -z notext is given..

This LGTM, I had a similar change in my repo while trying to fix PR35720.

Dec 23 2017, 6:43 PM

Dec 22 2017

emaste added a comment to D41541: [ELF] - Allow using PLT relocations when "-z notext" is given..

I can confirm this fixes the reduced test case, but the 'hello world' test on FreeBSD still fails, now with:

Dec 22 2017, 7:52 AM
emaste added inline comments to D41541: [ELF] - Allow using PLT relocations when "-z notext" is given..
Dec 22 2017, 6:28 AM

Dec 20 2017

emaste added a comment to D23934: Add a -ffixed-date-time= flag that sets the initial value of __DATE__, __TIME__, __TIMESTAMP__.

Ping?

Dec 20 2017, 12:04 PM

Dec 18 2017

emaste added a reviewer for D41207: Rename --icf-data and add a corresponding flag for functions: dim.
Dec 18 2017, 7:46 AM

Dec 13 2017

emaste added a comment to D41057: Avoid a gnu readelf warning.

FWIW our standard readelf in FreeBSD comes from the ELF Tool Chain project and doesn't have the warning. It's only when the GNU readelf package is installed that we'd encounter this.

Dec 13 2017, 12:55 PM

Dec 11 2017

emaste added inline comments to D41057: Avoid a gnu readelf warning.
Dec 11 2017, 5:31 PM

Dec 10 2017

emaste accepted D41037: Create reserved symbols early so they can be versioned.

This LGTM (but @ruiu or @grimar please review as well).

Dec 10 2017, 6:01 PM
emaste added inline comments to D41057: Avoid a gnu readelf warning.
Dec 10 2017, 6:00 PM
emaste added a comment to D41037: Create reserved symbols early so they can be versioned.

From e-mail followup @dim linked against libc linked with pre-patch lld.

Dec 10 2017, 5:46 PM

Dec 9 2017

emaste added a comment to D41037: Create reserved symbols early so they can be versioned.

I successfully linked all of FreeBSD/amd64 with LLD head + this patch.

Dec 9 2017, 5:31 AM

Dec 8 2017

emaste added a comment to D41037: Create reserved symbols early so they can be versioned.

Spoke too soon - buildworld is actually failing elsewhere now, with my backported patch in lld.

/usr/obj/usr/home/emaste/src/freebsd/amd64.amd64/tmp/usr/bin/ld: error: relocation R_X86_64_PLT32 cannot refer to absolute symbol: __tls_get_addr
>>> defined in <internal>
>>> referenced by runetype.h:98 (/usr/obj/usr/home/emaste/src/freebsd/amd64.amd64/tmp/usr/include/runetype.h:98)
>>>               xlat16_iconv.pico:(kiconv_xlat16_open)

I probably have an error in my backport attempt though, this other library (lib/libkiconv) linked succesfully with lld head + this patch.

Dec 8 2017, 6:40 PM
emaste added a comment to D41037: Create reserved symbols early so they can be versioned.

I tested this patch against lld head and confirm it resolves the issue via my reproducer. I also backported it to lld 5.0.1 (the version in FreeBSD HEAD) , rebuilt, and confirm that sbrk() is functional.

Dec 8 2017, 5:55 PM

Nov 20 2017

emaste added a reviewer for D32172: Port asan to FreeBSD AArch64: kcc.
Nov 20 2017, 6:27 PM
emaste added a reviewer for D32175: Add the llvm asan support: kcc.
Nov 20 2017, 6:27 PM

Nov 4 2017

emaste added inline comments to D39464: Define fs::allocate_file which preallocates disk blocks..
Nov 4 2017, 6:25 PM

Oct 31 2017

emaste added a comment to D39459: Rename SymbolBody -> Symbol..

Thanks for including the steps you used to effect the mechanical change! Having maintained long-lived derived branches of open source projects in the past I've found that sort of thing very helpful.

Oct 31 2017, 10:41 AM

Oct 14 2017

emaste committed rL315814: libunwind: document tested FreeBSD configs and sort OS list.
libunwind: document tested FreeBSD configs and sort OS list
Oct 14 2017, 10:04 AM
emaste closed D38900: libunwind: document tested FreeBSD configs and sort OS list by committing rL315814: libunwind: document tested FreeBSD configs and sort OS list.
Oct 14 2017, 10:04 AM

Oct 13 2017

emaste updated subscribers of D38900: libunwind: document tested FreeBSD configs and sort OS list.

@joerg might have insight on ppc, sparc64, arm on NetBSD.

Oct 13 2017, 2:01 PM
emaste created D38900: libunwind: document tested FreeBSD configs and sort OS list.
Oct 13 2017, 12:36 PM

Oct 6 2017

emaste added a comment to D38142: FreeBSD kernel debugging fixes.

Since I'm a first-time contributor, I definitely don't have commit privileges. Should I apply for them, or is someone else supposed to commit this for me?

Oct 6 2017, 12:09 PM

Oct 2 2017

emaste committed rL314672: Improve FreeBSD kernel debugging.
Improve FreeBSD kernel debugging
Oct 2 2017, 7:37 AM
emaste closed D38142: FreeBSD kernel debugging fixes by committing rL314672: Improve FreeBSD kernel debugging.
Oct 2 2017, 7:36 AM

Sep 30 2017

emaste added a comment to D38142: FreeBSD kernel debugging fixes.

I realized I don't have my LLVM SVN creds on my laptop, so will commit this when I return home in a couple of days if it's still not done.

Sep 30 2017, 12:28 AM

Sep 27 2017

emaste added a comment to D38142: FreeBSD kernel debugging fixes.

Since I'm a first-time contributor, I definitely don't have commit privileges. Should I apply for them, or is someone else supposed to commit this for me?

Sep 27 2017, 11:08 AM

Sep 16 2017

emaste added a comment to D34689: [LLD][ELF] Pre-create ThunkSections at Target specific intervals.

Very small nit noticed on another look: there are couple of instances of Mib that should be MiB

Sep 16 2017, 5:04 AM

Sep 13 2017

emaste accepted D37825: Accept not only --foo but also -foo..
Sep 13 2017, 6:21 PM

Sep 2 2017

emaste committed rL312431: Add test case for attach-by-pid from different cwd.
Add test case for attach-by-pid from different cwd
Sep 2 2017, 6:48 PM
emaste closed D32522: Test case for the issue raised in D32271 by committing rL312431: Add test case for attach-by-pid from different cwd.
Sep 2 2017, 6:48 PM
emaste committed rL312430: FreeBSD: attach to pid from different cwd.
FreeBSD: attach to pid from different cwd
Sep 2 2017, 6:43 PM
emaste closed D32271: Patch to Attach pid successfully from different dir by committing rL312430: FreeBSD: attach to pid from different cwd.
Sep 2 2017, 6:43 PM
emaste updated the diff for D32271: Patch to Attach pid successfully from different dir.

correct variable

Sep 2 2017, 4:32 PM
emaste added inline comments to D32522: Test case for the issue raised in D32271.
Sep 2 2017, 2:33 PM
emaste added a comment to D32522: Test case for the issue raised in D32271.

Testing this just now I got:

FAIL: LLDB (/usr/bin/cc-x86_64) :: test_attach_to_process_from_different_dir_by_id (TestProcessAttach.ProcessAttachTestCase)
======================================================================          
ERROR: test_attach_to_process_from_different_dir_by_id (TestProcessAttach.ProcessAttachTestCase)
   Test attach by process id                                                    
----------------------------------------------------------------------          
Traceback (most recent call last):                                              
  File "/tank/emaste/src/llvm/tools/lldb/packages/Python/lldbsuite/test/functionalities/process_attach/TestProcessAttach.py", line 44, in test_attach_to_process_from_different_dir_by_id
    os.mkdir(os.path.join(os.getcwd(),'newdir'))                                
OSError: [Errno 17] File exists: '/tank/emaste/src/llvm/tools/lldb/packages/Python/lldbsuite/test/functionalities/process_attach/newdir'
Config=x86_64-/usr/bin/cc

investigating...

Sep 2 2017, 10:09 AM

Aug 25 2017

emaste added a comment to D14700: [mips][freebsd] Explicitly select PC-relative Personality and LSDA encodings..

See D23113

Aug 25 2017, 2:40 PM

Aug 18 2017

emaste added a comment to D13104: Mips - Mark the section .eh_frame as writeable for pic.

We also face this problem on FreeBSD/mips64 when trying to link with lld.

Aug 18 2017, 8:47 AM
emaste added a comment to D34689: [LLD][ELF] Pre-create ThunkSections at Target specific intervals.

(Aside, for FreeBSD we are rather eager to have this support arrive in lld upstream and will from there backport it to the lld 5.0.0 in our src tree. As it stands we cannot link the full base system with lld because lldb has grown too large -- https://bugs.freebsd.org/215691)

Aug 18 2017, 6:25 AM

Aug 17 2017

emaste added inline comments to D34689: [LLD][ELF] Pre-create ThunkSections at Target specific intervals.
Aug 17 2017, 5:23 PM
emaste added a comment to D36823: [LLD][ELF] Read ARM BuildAttributes section to determine supported features..

Thank you, this will be quite useful for FreeBSD and the change seems reasonable to me.

Aug 17 2017, 8:32 AM

Aug 16 2017

emaste added inline comments to D36812: Add support for writing 64-bit symbol tables for archives when offsets become too large for 32-bit.
Aug 16 2017, 5:49 PM
emaste committed rL311003: convert hard tabs to spaces in PlatformOpenBSD.cpp.
convert hard tabs to spaces in PlatformOpenBSD.cpp
Aug 16 2017, 6:12 AM
emaste committed rL311002: Make i386-*-freebsd expression work on JIT path.
Make i386-*-freebsd expression work on JIT path
Aug 16 2017, 5:57 AM
emaste closed D34776: Make i386-*-freebsd expression work on JIT path by committing rL311002: Make i386-*-freebsd expression work on JIT path.
Aug 16 2017, 5:57 AM

Aug 13 2017

emaste added a comment to D32522: Test case for the issue raised in D32271.

Ping for two open questions

Aug 13 2017, 6:08 AM

Aug 10 2017

emaste committed rL310644: remove FreeBSD xfail decorator from TestCppNsImport.
remove FreeBSD xfail decorator from TestCppNsImport
Aug 10 2017, 11:27 AM
emaste committed rL310626: remove FreeBSD xfail decorator from TestCallStdStringFunction.
remove FreeBSD xfail decorator from TestCallStdStringFunction
Aug 10 2017, 10:02 AM
emaste committed rL310624: remove FreeBSD xfail decorator from TestExitDuringBreak.
remove FreeBSD xfail decorator from TestExitDuringBreak
Aug 10 2017, 9:49 AM
emaste committed rL310591: Report inferior signals as signals, not exceptions, on FreeBSD.
Report inferior signals as signals, not exceptions, on FreeBSD
Aug 10 2017, 6:48 AM
emaste closed D35223: Report inferior SIGSEGV/SIGILL/SIGBUS/SIGFPE as a signal instead of an exception on freebsd by committing rL310591: Report inferior signals as signals, not exceptions, on FreeBSD.
Aug 10 2017, 6:48 AM

Aug 9 2017

emaste accepted D35223: Report inferior SIGSEGV/SIGILL/SIGBUS/SIGFPE as a signal instead of an exception on freebsd.

Still would appreciate diffs uploaded with full context - i.e.,

Note that you can upload patches created through various diff tools, including git and svn. To make reviews easier, please always include as much context as possible with your diff! Don’t worry, Phabricator will automatically send a diff with a smaller context in the review email, but having the full file in the web interface will help the reviewer understand your code.
Aug 9 2017, 12:34 PM
emaste added a comment to D36538: Remove RISCV from LLVM_ALL_TARGETS in CMakeLists.txt.

Adding the comment above the list seems like a good idea to me.

Aug 9 2017, 12:17 PM
emaste added a comment to D34776: Make i386-*-freebsd expression work on JIT path.

I've committed this to FreeBSD's copy of lldb in r322326. @labath if you're happy with this patch I will commit to lldb for @karnajitw. I'm not sure how the patch ended up with a conflict, but it's just a whitespace issue.

Aug 9 2017, 12:13 PM

Aug 8 2017

emaste added a reviewer for D34776: Make i386-*-freebsd expression work on JIT path: labath.
Aug 8 2017, 11:54 AM
emaste accepted D34776: Make i386-*-freebsd expression work on JIT path.
Aug 8 2017, 11:54 AM
emaste added a comment to D35223: Report inferior SIGSEGV/SIGILL/SIGBUS/SIGFPE as a signal instead of an exception on freebsd.

Actually, I think you probably need to extend the @skipIfLinux to apply to freebsd as well.

Aug 8 2017, 10:27 AM

Aug 7 2017

emaste added a comment to D34776: Make i386-*-freebsd expression work on JIT path.

With this patch i386 test results are comparable to amd64, and I'm happy with it from a FreeBSD perspective (modulo the PlatformOpenBSD patch conflict).

Aug 7 2017, 6:27 PM
emaste added a comment to D34776: Make i386-*-freebsd expression work on JIT path.

The change in PlatformOpenBSD.cpp failed to apply for me (although it was trivial to manually apply it).

Aug 7 2017, 12:34 PM