Page MenuHomePhabricator

richard.barton.arm (Richard Barton)
Staff Engineer - LLVM

Projects

User does not belong to any projects.

User Details

User Since
Oct 3 2013, 2:16 AM (333 w, 1 d)

Recent Activity

Mon, Feb 17

richard.barton.arm added inline comments to D73951: [Clang] [Driver]Add logic to search for flang frontend.
Mon, Feb 17, 12:56 AM · Restricted Project

Fri, Feb 14

richard.barton.arm added inline comments to D73951: [Clang] [Driver]Add logic to search for flang frontend.
Fri, Feb 14, 8:50 AM · Restricted Project
richard.barton.arm added a comment to D73951: [Clang] [Driver]Add logic to search for flang frontend.

Still not sure why all the tests are needed here. The first one looks fine, i.e. we test that --fortran-fe=<path to copy of driver binary we just created> calls to that copy. The second one appears to test the default behaviour with no option, but why does it do it via a different name for clang, why would that make a difference? The second uses a different name for both driver and frontend. Why is that test relevant to the change you have made? As far as I can see, the name of the driver binary has no bearing on its behaviour wrt. calling a frontend binary. What am I missing?

Fri, Feb 14, 8:41 AM · Restricted Project
richard.barton.arm added a comment to D73951: [Clang] [Driver]Add logic to search for flang frontend.

I'd still like to see the nits addressed and comments on the tests addressed before approving.

Fri, Feb 14, 1:36 AM · Restricted Project

Wed, Feb 5

richard.barton.arm added inline comments to D73951: [Clang] [Driver]Add logic to search for flang frontend.
Wed, Feb 5, 6:48 AM · Restricted Project

Tue, Feb 4

richard.barton.arm requested changes to D73951: [Clang] [Driver]Add logic to search for flang frontend.

A few comments running through that need addressing.

Tue, Feb 4, 10:59 AM · Restricted Project

Sep 23 2019

richard.barton.arm added a comment to D63607: [clang][driver] Add basic --driver-mode=flang support for fortran.

I think the behaviour for missing flang is fine for now, and I think we can improve on it later on.
We ought to codify (if it is not done already) where flang looks for tools to exec, because I think PATH is probably not the only place it could look to (directory of clang binary, other relative paths, other environment variables and commandline options, etc.) The new test will need revising once we get there.

Sep 23 2019, 12:45 AM · Restricted Project

Sep 18 2019

richard.barton.arm added a comment to D63607: [clang][driver] Add basic --driver-mode=flang support for fortran.

Thanks for the updates. I think this is now looking good and matches the RFC already posted.

Sep 18 2019, 10:16 AM · Restricted Project

Sep 17 2019

richard.barton.arm added a comment to D63607: [clang][driver] Add basic --driver-mode=flang support for fortran.

Hi Peter

Sep 17 2019, 11:11 AM · Restricted Project

Nov 21 2018

richard.barton.arm added a comment to D53210: Revert 344389 "Revert r344375 "[Driver] check for exit code from SIGPIPE"".

Hi @nickdesaulniers - thanks for the clarification. I was suffering from some PEBCAK of my own when I thought the commits were not on master. Thanks for these patches - a great help.

Nov 21 2018, 11:50 AM

Nov 20 2018

richard.barton.arm added a comment to D53210: Revert 344389 "Revert r344375 "[Driver] check for exit code from SIGPIPE"".

Hi @nickdesaulniers
I have run into this too recently so would love to see this patch land. Did you get anywhere with those lld test failures?
Ta
Rich

Nov 20 2018, 11:29 AM

Aug 8 2017

richard.barton.arm added a comment to D35569: [ARM] Remove FeatureNoARM implies ModeThumb..

Agreed with what Florian says. My understanding of -target is that it is used to set up which flavour of toolchain we are using and using it to set the architecture is a shorthand for saying '-target=arm... -march=armvX'. I also expect it to be somewhat like GCC where the 'arm-...' triples cover ARM/AArch32 architecture in general and not the ARM ISA per se.

Aug 8 2017, 6:44 AM

Feb 9 2017

richard.barton.arm accepted D29472: [ARM] Add support for armv7ve triple in llvm (PR31358)..

Thanks Manoj - this now also LGTM.

Feb 9 2017, 1:11 PM
richard.barton.arm accepted D29773: Add support for armv7ve flag in clang (PR31358)..

This all LGTM. If I can assume I am allow to approve, then I approve.

Feb 9 2017, 11:50 AM
richard.barton.arm requested changes to D29472: [ARM] Add support for armv7ve triple in llvm (PR31358)..

Hi Manoj

Feb 9 2017, 11:38 AM

Feb 3 2017

richard.barton.arm requested changes to D29472: [ARM] Add support for armv7ve triple in llvm (PR31358)..

Hi Manoj

Feb 3 2017, 11:41 AM

Aug 18 2016

richard.barton.arm added a comment to D23500: [ARM] Correct ARMv8*-A optional extension definitions in TargetParser.

Hi all - I was puzzling over how to write a new unit test that would catch this error without just copying the structure of the target parser itself. The best I could come up with is to define a for each architecture extension, a set of all architectures that must enable it by default. You could then iterate over all the extensions and for each check all the architectures and test that those in the set have the enum set and those not in the set do not.

Aug 18 2016, 6:55 AM

Aug 16 2016

richard.barton.arm added reviewers for D23500: [ARM] Correct ARMv8*-A optional extension definitions in TargetParser: t.p.northover, rengolin.
Aug 16 2016, 7:06 AM

Aug 15 2016

richard.barton.arm added a comment to D21428: Fix feature RAS assigned to ARMv8.1 when it should be ARMv8.2.

These macros are available for setting compiler-provided preprocessing macros like ACLE. We are not using any of the ARMv8.2-A ones because the ACLE has not caught up sufficiently with the architecture to have defined any.

Aug 15 2016, 11:10 AM
richard.barton.arm retitled D23500: [ARM] Correct ARMv8*-A optional extension definitions in TargetParser from to [ARM] Correct ARMv8*-A optional extension definitions in TargetParser.
Aug 15 2016, 4:32 AM

Jul 13 2016

richard.barton.arm updated subscribers of D22254: [ARM, AArch64] Adds builtins to get current SP and PC.
Jul 13 2016, 10:36 AM
richard.barton.arm added a comment to D22272: ARM: define __ARM_VFPV5__ when present..

Hi all

Jul 13 2016, 12:55 AM

Mar 23 2016

richard.barton.arm added a comment to D18347: [PATCH] Fix thread_annotation negtest for thread safety..

Hi Eric - no problem at all, you practically wrote it for me after all ;-)

Mar 23 2016, 2:09 PM
richard.barton.arm updated the diff for D18347: [PATCH] Fix thread_annotation negtest for thread safety..

My local run with no threads works with this updated patch.

Mar 23 2016, 6:37 AM

Mar 22 2016

richard.barton.arm added a comment to D18347: [PATCH] Fix thread_annotation negtest for thread safety..

Thanks for the help Eric. I'm just running a new patch now will put the new patch up when I get back in to office tomorrow.

Mar 22 2016, 3:20 PM
richard.barton.arm added a comment to D18347: [PATCH] Fix thread_annotation negtest for thread safety..

Hi Eric

Mar 22 2016, 1:55 PM
richard.barton.arm updated the diff for D18347: [PATCH] Fix thread_annotation negtest for thread safety..

Sorry - not sure what happened there. That looks better for me.

Mar 22 2016, 11:59 AM
richard.barton.arm retitled D18347: [PATCH] Fix thread_annotation negtest for thread safety. from to [PATCH] Fix thread_annotation negtest for thread safety..
Mar 22 2016, 4:53 AM

Mar 3 2016

richard.barton.arm added a comment to D15283: [ARMv8-M] Add Clang targeting for ARMv8-M Baseline/Mainline.

Hi Bradley

Mar 3 2016, 4:40 AM

Feb 18 2016

richard.barton.arm closed D13498: Silence spurious warning from lit about missing clang-interpreter.
Feb 18 2016, 9:35 AM

Oct 21 2015

richard.barton.arm closed D13930: Fix __ARM_FP value for sp-only FPUs with Half-precision.

Committed as r250888

Oct 21 2015, 3:07 AM
richard.barton.arm retitled D13930: Fix __ARM_FP value for sp-only FPUs with Half-precision from to Fix __ARM_FP value for sp-only FPUs with Half-precision.
Oct 21 2015, 2:15 AM

Oct 16 2015

richard.barton.arm closed D13500: Improve helpfulness of assert message when number of diagnostics overflows number of available enum values..

Committed as r250546 with the suggested change.

Oct 16 2015, 1:17 PM

Oct 9 2015

richard.barton.arm added a reviewer for D13500: Improve helpfulness of assert message when number of diagnostics overflows number of available enum values.: rsmith.
Oct 9 2015, 9:51 AM

Oct 7 2015

richard.barton.arm abandoned D13495: Improve the helpfulness of an assert message.

Abandoning as cfe-commits was not on subscribers

Oct 7 2015, 1:31 AM
richard.barton.arm retitled D13500: Improve helpfulness of assert message when number of diagnostics overflows number of available enum values. from to Improve helpfulness of assert message when number of diagnostics overflows number of available enum values..
Oct 7 2015, 1:30 AM
richard.barton.arm abandoned D13496: Convert test to take input from stdin rather than reading a file.

Abandoning as I forgot to add cfe-commits.

Oct 7 2015, 1:29 AM
richard.barton.arm retitled D13499: Convert test to take input from stdin rather than reading a file from to Convert test to take input from stdin rather than reading a file.
Oct 7 2015, 1:28 AM
richard.barton.arm retitled D13498: Silence spurious warning from lit about missing clang-interpreter from to Silence spurious warning from lit about missing clang-interpreter.
Oct 7 2015, 1:27 AM
richard.barton.arm abandoned D13497: Silence warning from lit about missing clang-interpreter.

Adandoning and starting again with cfe-commits as subscriber.

Oct 7 2015, 1:23 AM
richard.barton.arm removed a reviewer for D13497: Silence warning from lit about missing clang-interpreter: cfe-commits.
Oct 7 2015, 1:16 AM
richard.barton.arm removed a reviewer for D13496: Convert test to take input from stdin rather than reading a file: cfe-commits.
Oct 7 2015, 1:16 AM
richard.barton.arm added a reviewer for D13495: Improve the helpfulness of an assert message: rafael.
Oct 7 2015, 12:59 AM
richard.barton.arm added a reviewer for D13496: Convert test to take input from stdin rather than reading a file: cfe-commits.
Oct 7 2015, 12:52 AM
richard.barton.arm added a reviewer for D13497: Silence warning from lit about missing clang-interpreter: cfe-commits.
Oct 7 2015, 12:52 AM
richard.barton.arm retitled D13497: Silence warning from lit about missing clang-interpreter from to Silence warning from lit about missing clang-interpreter.
Oct 7 2015, 12:51 AM
richard.barton.arm retitled D13496: Convert test to take input from stdin rather than reading a file from to Convert test to take input from stdin rather than reading a file.
Oct 7 2015, 12:46 AM
richard.barton.arm updated subscribers of D13495: Improve the helpfulness of an assert message.
Oct 7 2015, 12:37 AM
richard.barton.arm retitled D13495: Improve the helpfulness of an assert message from to Improve the helpfulness of an assert message.
Oct 7 2015, 12:35 AM

Oct 2 2015

richard.barton.arm added a comment to D12633: Implement ACLE 2.0 macros of chapters 6.6 and 6.7 for [ARM] and [Aarch64] targets.

What do you think about this Renato? Although Alexandros has added a LangOption which is strictly speaking to do with CodeGen rather than Lang, I think this is the only pragmatica way to do this without needing re-engineering work to expose the codegen options to this phase in clang.

Oct 2 2015, 7:18 AM

Oct 1 2015

richard.barton.arm added inline comments to D13298: Fix PR23871: Passing a string literal to .byte directive crashes the assembler.
Oct 1 2015, 2:02 AM

Sep 28 2015

richard.barton.arm added inline comments to D12937: [ARM] Handle +t2dsp feature as an ArchExtKind in ARMTargetParser.def.
Sep 28 2015, 2:00 AM

Sep 25 2015

richard.barton.arm added inline comments to D12937: [ARM] Handle +t2dsp feature as an ArchExtKind in ARMTargetParser.def.
Sep 25 2015, 5:11 AM

Sep 24 2015

richard.barton.arm added a comment to D12633: Implement ACLE 2.0 macros of chapters 6.6 and 6.7 for [ARM] and [Aarch64] targets.

In the overall approach, whilst this is the pragmatic approach to get the job done, it clearly takes liberties with the definition of LangOpt as this your new flag is most certainly a CodeGenOpt rather than a LangOpt. I'd like to see what the others say as I don't see a better solution.

Sep 24 2015, 8:30 AM
richard.barton.arm added inline comments to D12633: Implement ACLE 2.0 macros of chapters 6.6 and 6.7 for [ARM] and [Aarch64] targets.
Sep 24 2015, 8:01 AM

Sep 21 2015

richard.barton.arm added a comment to D12633: Implement ACLE 2.0 macros of chapters 6.6 and 6.7 for [ARM] and [Aarch64] targets.

Yes - I would be interested in knowing the rationale behind the original AArch64 setting of __ARM_FP_FPENV_ROUNDING.

Sep 21 2015, 9:57 AM

Sep 17 2015

richard.barton.arm added a comment to D12633: Implement ACLE 2.0 macros of chapters 6.6 and 6.7 for [ARM] and [Aarch64] targets.

Hi all

Sep 17 2015, 5:43 AM

Sep 8 2015

richard.barton.arm added a comment to D12692: LLVM does not distinguish Cortex-M4 from Cortex-M4F neither Cortex-R5 from R5F.

Hi Alexandros

Sep 8 2015, 2:58 PM

Jul 31 2015

richard.barton.arm added a comment to D11590: Arch extensions and default target features in TargetParser.

Agree with Renato's comments, apart from one.

Jul 31 2015, 9:19 AM
richard.barton.arm added inline comments to D11640: [ARM] Proper handling generic cpus.
Jul 31 2015, 4:12 AM
richard.barton.arm added inline comments to D11590: Arch extensions and default target features in TargetParser.
Jul 31 2015, 3:35 AM
richard.barton.arm added a comment to D11590: Arch extensions and default target features in TargetParser.

Regarding the table, apart from the above Divide concern, I notice also

  • the lack of MP extension for Cortex-R5

Are we sure Cortex-R5 has MP extensions?

Jul 31 2015, 3:34 AM
richard.barton.arm added inline comments to D11640: [ARM] Proper handling generic cpus.
Jul 31 2015, 12:59 AM
richard.barton.arm added inline comments to D11590: Arch extensions and default target features in TargetParser.
Jul 31 2015, 12:56 AM

Jul 30 2015

richard.barton.arm added a comment to D11590: Arch extensions and default target features in TargetParser.

Comments inline.

Jul 30 2015, 11:48 AM
richard.barton.arm added a comment to D11590: Arch extensions and default target features in TargetParser.

Agreed - I can look at the next revision from Alexandros regarding adding MP and Virtualization to this table (I am less fussed about OS)

Jul 30 2015, 5:10 AM
richard.barton.arm added a comment to D11590: Arch extensions and default target features in TargetParser.

Thanks for explaining the Krait situation - this now makes sense to me and I agree that it is not ideal.

Jul 30 2015, 4:28 AM
richard.barton.arm added a comment to D11590: Arch extensions and default target features in TargetParser.

Alexandros asked if I could look over the table of extensions, but I have some points on the code in general.

Jul 30 2015, 4:08 AM

Jul 14 2015

richard.barton.arm added a comment to D11100: API for retrieving default FPU target features from TargetParser.

Be careful - you are linking to the internal infocentre rather than the public one.

Jul 14 2015, 5:50 AM

Jul 3 2015

richard.barton.arm added a comment to D10897: Reworked the test part of r241149 "Fix PR23872: Integrated assembler error message when using .type directive with @ in AArch32 assembly.".

Apologies for the "miss" in the original review gents.

Jul 3 2015, 3:30 AM

Jul 1 2015

richard.barton.arm added a comment to D10651: PR23872 Integrated assembler error message when using .type directive with @ in AArch32 assembly.

LGTM

Jul 1 2015, 1:37 AM

Jun 25 2015

richard.barton.arm added a comment to D10651: PR23872 Integrated assembler error message when using .type directive with @ in AArch32 assembly.

Wont the definition getAllowAtInIdentifier tell you?

Jun 25 2015, 8:19 AM

Jun 24 2015

richard.barton.arm added a comment to D10651: PR23872 Integrated assembler error message when using .type directive with @ in AArch32 assembly.

Are there any other targets that we should test it on?

Jun 24 2015, 11:28 AM
richard.barton.arm added a comment to D10651: PR23872 Integrated assembler error message when using .type directive with @ in AArch32 assembly.

Hi Gabor

Jun 24 2015, 11:26 AM

Jun 18 2015

richard.barton.arm added a comment to D10535: Allow case-insensitive values for -march for AArch64 target in line with GCC..

Again no test that the case insensitivity also applies to the feature extensions. Other than that, this seems the same as the previous patch about -mcpu and I see no reason not to commit it.

Jun 18 2015, 7:02 AM

Jun 12 2015

richard.barton.arm added a comment to D10412: Allow case-insensitive values for -mcpu for AArch64 target in line with GCC..

I have one comment about a missing regression test, but otherwise this patch seems fine to me. I understand the previous version broke the ARM bot, so I will let Renato have last say.

Jun 12 2015, 9:21 AM

Jun 1 2015

richard.barton.arm added a comment to D10162: Allow case-insensitive values for -mcpu for ARM in line with GCC..

My only observation is that you have changed the Krait codepath, but have not added a test for that.

Jun 1 2015, 10:02 AM

May 8 2015

richard.barton.arm added a comment to D8501: [AArch64] Add v8.1a atomic instructions.

Hi Tim

May 8 2015, 3:32 AM

May 5 2015

richard.barton.arm added a comment to D9476: [PATCH] Allow case-insensitive values for -mcpu for ARM and AArch64 targets in line with GCC..

Hi Gabor, Joerg

May 5 2015, 8:03 AM

Feb 27 2015

richard.barton.arm added a comment to D7730: Make the driver accept all four variants of the target option.

Hi Gabor

Feb 27 2015, 2:53 AM

Feb 19 2015

richard.barton.arm added a comment to D7730: Make the driver accept all four variants of the target option.

There's a consensus in the Unix world that only two such syntaxes should be
supported:

-key value
--key=value

We only accept those two, and not the other two. This is intentional.

Feb 19 2015, 4:44 AM
richard.barton.arm added a comment to D7730: Make the driver accept all four variants of the target option.

I think that having -target=<triple> and --target <triple> only is awkward.
Why have two version of the same option with different rules on whether they
accept = or spaced arguments?

Feb 19 2015, 3:47 AM

Oct 10 2014

richard.barton.arm added a comment to D5721: [AArch64] Add workaround for Cortex-A53 erratum (835769).

Hi Bradley

Oct 10 2014, 10:34 AM

Sep 1 2014

richard.barton.arm updated subscribers of D4870: Segfault in AArch64 backend with -g and -mbig-endian.
Sep 1 2014, 3:24 PM