chandlerc (Chandler Carruth)Administrator
User

Projects

User does not belong to any projects.

User Details

User Since
Jul 7 2012, 2:54 PM (327 w, 5 d)
Roles
Administrator

Recent Activity

Yesterday

chandlerc committed rL344769: [TI removal] Remove `TerminatorInst` from the IR type system!.
[TI removal] Remove `TerminatorInst` from the IR type system!
Thu, Oct 18, 5:24 PM
chandlerc committed rL344768: [TI removal] Switch some newly added code over to use `Instruction`.
[TI removal] Switch some newly added code over to use `Instruction`
Thu, Oct 18, 5:24 PM
chandlerc committed rL344764: [TI removal] Update the C API for the move away from `TerminatorInst`..
[TI removal] Update the C API for the move away from `TerminatorInst`.
Thu, Oct 18, 4:06 PM
chandlerc closed D52968: [TI removal] Update the C API for the move away from `TerminatorInst`..
Thu, Oct 18, 4:06 PM
chandlerc updated the diff for D52968: [TI removal] Update the C API for the move away from `TerminatorInst`..

Fix one more place.

Thu, Oct 18, 3:47 PM
chandlerc updated the diff for D52968: [TI removal] Update the C API for the move away from `TerminatorInst`..

Update based on Jame's suggestion.

Thu, Oct 18, 3:47 PM
chandlerc added a comment to D52968: [TI removal] Update the C API for the move away from `TerminatorInst`..

Done.

Thu, Oct 18, 3:47 PM
chandlerc updated the diff for D52968: [TI removal] Update the C API for the move away from `TerminatorInst`..

Ping, rebase, and update to remove all but the type system bits to match the
progress outside of the C API.

Thu, Oct 18, 2:39 AM
chandlerc committed rC344730: [TI removal] Test predicate rather than casting to detect a terminator.
[TI removal] Test predicate rather than casting to detect a terminator
Thu, Oct 18, 1:19 AM
chandlerc committed rL344730: [TI removal] Test predicate rather than casting to detect a terminator.
[TI removal] Test predicate rather than casting to detect a terminator
Thu, Oct 18, 1:18 AM
chandlerc committed rL344728: [TI removal] Generically discuss terminators rather than use the soon to.
[TI removal] Generically discuss terminators rather than use the soon to
Thu, Oct 18, 12:46 AM
chandlerc committed rL344727: [TI removal] Remove discussion of `TerminatorInst` from the LLVM.
[TI removal] Remove discussion of `TerminatorInst` from the LLVM
Thu, Oct 18, 12:42 AM
chandlerc committed rL344726: [TI removal] Remove TerminatorInst references from bindings..
[TI removal] Remove TerminatorInst references from bindings.
Thu, Oct 18, 12:42 AM

Wed, Oct 17

chandlerc committed rL344719: [TI removal] Switch simple loop unswitch to `Instruction`..
[TI removal] Switch simple loop unswitch to `Instruction`.
Wed, Oct 17, 5:42 PM
chandlerc committed rL344718: [TI removal] Switch NewGVN to directly use `Instruction`..
[TI removal] Switch NewGVN to directly use `Instruction`.
Wed, Oct 17, 5:41 PM
chandlerc committed rL344717: [TI removal] Use `Instruction` instead of `TerminatorInst` for.
[TI removal] Use `Instruction` instead of `TerminatorInst` for
Wed, Oct 17, 5:41 PM
chandlerc committed rL344716: [TI removal] Update CodeExtractor to use Instruction directly..
[TI removal] Update CodeExtractor to use Instruction directly.
Wed, Oct 17, 5:41 PM
chandlerc committed rL344715: [TI removal] Switch ObjCARC code to directly use the nice range-based.
[TI removal] Switch ObjCARC code to directly use the nice range-based
Wed, Oct 17, 5:40 PM
chandlerc committed rL344714: [TI removal] Switch MergeFunctions to directly use Instruction API..
[TI removal] Switch MergeFunctions to directly use Instruction API.
Wed, Oct 17, 5:40 PM
chandlerc committed rL344713: [TI removal] Switch an analysis to just use Instruction..
[TI removal] Switch an analysis to just use Instruction.
Wed, Oct 17, 5:38 PM
chandlerc committed rL344704: Update list of contacted companies..
Update list of contacted companies.
Wed, Oct 17, 11:55 AM
chandlerc updated the diff for D52968: [TI removal] Update the C API for the move away from `TerminatorInst`..

Ping again! Also rebase...

Wed, Oct 17, 11:05 AM

Tue, Oct 16

chandlerc committed rL344656: Major update to relicensing page..
Major update to relicensing page.
Tue, Oct 16, 4:30 PM
chandlerc closed D53269: Major update to relicensing page..
Tue, Oct 16, 4:30 PM

Mon, Oct 15

chandlerc committed rC344504: [TI removal] Make `getTerminator()` return a generic `Instruction`..
[TI removal] Make `getTerminator()` return a generic `Instruction`.
Mon, Oct 15, 3:45 AM
chandlerc committed rL344504: [TI removal] Make `getTerminator()` return a generic `Instruction`..
[TI removal] Make `getTerminator()` return a generic `Instruction`.
Mon, Oct 15, 3:44 AM
chandlerc accepted D53246: [NewPM] teach -passes= to emit meaningful error messages.

LGTM, this is a great improvement IMO!

Mon, Oct 15, 3:40 AM
chandlerc committed rL344503: [TI removal] Rework `InstVisitor` to support visiting instructions that.
[TI removal] Rework `InstVisitor` to support visiting instructions that
Mon, Oct 15, 3:13 AM
chandlerc committed rL344502: [TI removal] Make variables declared as `TerminatorInst` and initialized.
[TI removal] Make variables declared as `TerminatorInst` and initialized
Mon, Oct 15, 3:07 AM
chandlerc committed rL344501: [TI removal] Remove `TerminatorInst` from GVN.h and GVN.cpp..
[TI removal] Remove `TerminatorInst` from GVN.h and GVN.cpp.
Mon, Oct 15, 3:02 AM
chandlerc committed rL344500: [TI removal] Remove `TerminatorInst` from SparsePropagation.h and.
[TI removal] Remove `TerminatorInst` from SparsePropagation.h and
Mon, Oct 15, 2:49 AM
chandlerc committed rL344499: [TI removal] Remove a dead forward declaration of TerminatorInst. NFC..
[TI removal] Remove a dead forward declaration of TerminatorInst. NFC.
Mon, Oct 15, 2:36 AM
chandlerc committed rL344498: [TI removal] Remove `TerminatorInst` from BasicBlockUtils.h.
[TI removal] Remove `TerminatorInst` from BasicBlockUtils.h
Mon, Oct 15, 2:36 AM
chandlerc committed rL344497: [TI removal] Just use Instruction in the CFG printer code. NFC..
[TI removal] Just use Instruction in the CFG printer code. NFC.
Mon, Oct 15, 2:35 AM
chandlerc committed rL344495: [TI removal] Remove a unnecessary use of `TerminatorInst` from an IR.
[TI removal] Remove a unnecessary use of `TerminatorInst` from an IR
Mon, Oct 15, 2:19 AM
chandlerc committed rL344494: [TI removal] Remove TerminatorInst as an input parameter from all public.
[TI removal] Remove TerminatorInst as an input parameter from all public
Mon, Oct 15, 2:19 AM
chandlerc added a comment to D52968: [TI removal] Update the C API for the move away from `TerminatorInst`..

Ping? Would be really nice to get a review here.

Mon, Oct 15, 2:03 AM
chandlerc created D53269: Major update to relicensing page..
Mon, Oct 15, 12:12 AM
chandlerc committed rL344491: Add a company that has signed the corporate agreement..
Add a company that has signed the corporate agreement.
Mon, Oct 15, 12:06 AM

Thu, Oct 11

chandlerc committed rCTE344225: Fix the qualification of `IntrusiveRefCntPtr` to use `llvm::`..
Fix the qualification of `IntrusiveRefCntPtr` to use `llvm::`.
Thu, Oct 11, 1:09 AM
chandlerc committed rL344225: Fix the qualification of `IntrusiveRefCntPtr` to use `llvm::`..
Fix the qualification of `IntrusiveRefCntPtr` to use `llvm::`.
Thu, Oct 11, 1:07 AM

Wed, Oct 10

chandlerc added a comment to D52915: [PM] keeping history when original SCC split and then merge into itself in the same round of SCC update ..

Ok, also had some time to really think about the core problem and approach. Write-up below.

Wed, Oct 10, 4:07 PM
chandlerc added a comment to D52915: [PM] keeping history when original SCC split and then merge into itself in the same round of SCC update ..

Basic suggestions on test case...

Wed, Oct 10, 4:02 PM
chandlerc added a comment to D52915: [PM] keeping history when original SCC split and then merge into itself in the same round of SCC update ..

Minor implementation improvement suggestion...

Wed, Oct 10, 3:59 PM

Mon, Oct 8

chandlerc committed rL344014: [CFG Printer] Add support for writing the dot files with a custom.
[CFG Printer] Add support for writing the dot files with a custom
Mon, Oct 8, 9:32 PM

Sun, Oct 7

chandlerc added a comment to D52968: [TI removal] Update the C API for the move away from `TerminatorInst`..

Sending this for pre-commit review as I want to be sure I'm doing the right thing w/ the C-API here.

Sun, Oct 7, 2:05 AM
chandlerc created D52968: [TI removal] Update the C API for the move away from `TerminatorInst`..
Sun, Oct 7, 2:05 AM
chandlerc committed rL343932: Add a listing of companies that have agreed to the relicensing..
Add a listing of companies that have agreed to the relicensing.
Sun, Oct 7, 12:08 AM
chandlerc closed D52960: Add a listing of companies that have agreed to the relicensing..
Sun, Oct 7, 12:08 AM

Fri, Oct 5

chandlerc created D52960: Add a listing of companies that have agreed to the relicensing..
Fri, Oct 5, 10:43 PM
chandlerc closed D52929: Update to relicensing website with developments over the past few months..
Fri, Oct 5, 10:42 PM
chandlerc committed rL343910: Update the relicensing page with the latest information..
Update the relicensing page with the latest information.
Fri, Oct 5, 9:51 PM
chandlerc accepted D51276: [New PM][PassTiming] implement -time-passes for the new pass manager.

Somewhat minor code simplification below. If it makes sense, feel free to land. If it doesn't honestly, feel free to explain why not and land. Even if I convince you, it's super simple to land as a follow-up. =]

Fri, Oct 5, 3:20 AM
chandlerc created D52929: Update to relicensing website with developments over the past few months..
Fri, Oct 5, 2:59 AM
chandlerc added a comment to D52915: [PM] keeping history when original SCC split and then merge into itself in the same round of SCC update ..

Just wanted to send a quick note that I saw this and am thinking about the approach and the problem. Thanks for looking at this (really tricky) problem!

Fri, Oct 5, 1:11 AM

Thu, Oct 4

chandlerc added a comment to D51276: [New PM][PassTiming] implement -time-passes for the new pass manager.

as for the per-pass-name aggregation - I was going to supply a follow-up patch that changes -time-passes from boolean option to -time-passes=<aggregation-type>.
Perhaps aggregation-type should be 'aggregate' by default, and 'invocations' providing per-invocation info.
Yet I dont want to just remove the possibility of getting a more fine-grained data that we have in this patch.

Thu, Oct 4, 3:38 AM
chandlerc added a comment to D51276: [New PM][PassTiming] implement -time-passes for the new pass manager.

Some more minor code comments inline...

Thu, Oct 4, 1:24 AM

Wed, Sep 19

chandlerc committed rL342588: Workaround a limitation of llvm::Any when used with types that have.
Workaround a limitation of llvm::Any when used with types that have
Wed, Sep 19, 1:46 PM

Sep 18 2018

chandlerc accepted D47858: [New PM] Introducing PassInstrumentation framework.

LGTM (provided Philip is still happy here, I know I asked to remove his plugin hooks).

Sep 18 2018, 4:33 PM
chandlerc added inline comments to D47858: [New PM] Introducing PassInstrumentation framework.
Sep 18 2018, 1:50 AM

Sep 15 2018

chandlerc added a comment to D47858: [New PM] Introducing PassInstrumentation framework.

It gets better and better. Another round of comments.

Sep 15 2018, 3:09 AM

Sep 13 2018

chandlerc added a comment to D50858: [M680x0] Add ELF and Triple info.

Before we add yet another experimental backend, I think it is somewhat important to know what the long-term user story is and the long-term maintenance story.

Sep 13 2018, 5:32 PM
chandlerc added inline comments to D47858: [New PM] Introducing PassInstrumentation framework.
Sep 13 2018, 5:11 AM
chandlerc added a comment to D47858: [New PM] Introducing PassInstrumentation framework.

This is looking really, really good. Some comments below.

Sep 13 2018, 3:17 AM

Sep 7 2018

chandlerc accepted D51768: [X86] Teach X86DAGToDAGISel::foldLoadStoreIntoMemOperand to handle loads in operand 1 of commutable operations..

LGTM, nice!

Sep 7 2018, 2:10 AM

Sep 4 2018

chandlerc committed rL341365: Revert r341269: [Constant Hoisting] Hoisting Constant GEP Expressions.
Revert r341269: [Constant Hoisting] Hoisting Constant GEP Expressions
Sep 4 2018, 6:40 AM
chandlerc committed rL341363: [x86/SLH] Add a real Clang flag and LLVM IR attribute for Speculative.
[x86/SLH] Add a real Clang flag and LLVM IR attribute for Speculative
Sep 4 2018, 5:39 AM
chandlerc committed rC341363: [x86/SLH] Add a real Clang flag and LLVM IR attribute for Speculative.
[x86/SLH] Add a real Clang flag and LLVM IR attribute for Speculative
Sep 4 2018, 5:39 AM
chandlerc closed D51157: [x86/SLH] Add a real Clang flag and LLVM IR attribute for Speculative Load Hardening..
Sep 4 2018, 5:39 AM
chandlerc added a comment to D51157: [x86/SLH] Add a real Clang flag and LLVM IR attribute for Speculative Load Hardening..

All outstanding comments addressed, and landing this. Thanks all for the reviews and let me know if I missed anything.

Sep 4 2018, 5:25 AM
chandlerc added a comment to D49678: [llvm-objcopy] Adding support for compressed DWARF debug sections..

Reverted temporarily in r341360 to unbreak my build and the aarch64 build bot. Let me know if I can help with the investigation in any way.

Sep 4 2018, 5:01 AM
chandlerc committed rL341360: Revert r341342: Dwarf .debug section compression support (zlib, zlib-gnu)..
Revert r341342: Dwarf .debug section compression support (zlib, zlib-gnu).
Sep 4 2018, 4:57 AM
chandlerc committed rL341358: [x86/SLH] Teach SLH to harden against the "ret2spec" attack by.
[x86/SLH] Teach SLH to harden against the "ret2spec" attack by
Sep 4 2018, 4:00 AM
chandlerc closed D50837: [x86/SLH] Teach SLH to harden against the "ret2spec" attack by implementing the proposed mitigation technique described in the original design document..
Sep 4 2018, 4:00 AM
chandlerc added a comment to D49678: [llvm-objcopy] Adding support for compressed DWARF debug sections..

There are bots obviously still failing so I'm going to revert....

Sep 4 2018, 3:53 AM
chandlerc committed rL341356: [x86/SLH] Teach SLH to harden indirect branches and switches without.
[x86/SLH] Teach SLH to harden indirect branches and switches without
Sep 4 2018, 3:47 AM
chandlerc closed D51083: [x86/SLH] Teach SLH to harden indirect branches and switches without retpolines..
Sep 4 2018, 3:47 AM
chandlerc added a comment to D51083: [x86/SLH] Teach SLH to harden indirect branches and switches without retpolines..

Thanks for the review, landing.

Sep 4 2018, 3:41 AM
chandlerc added a comment to D49678: [llvm-objcopy] Adding support for compressed DWARF debug sections..

This change broke several LLVM builders as well as our downstream Linux and macOS toolchain builders, would it be possible to either quickly fix the issue or revert it?

Have your builders picked up r341343? Some of the ubsan failures (and I suspect others) came from an unaligned memory write that I hadn't caught locally which is fixed now. If your builders still fail after this go ahead and revert and let me know. Sorry about this.

Sep 4 2018, 3:41 AM
chandlerc added a comment to D51525: [ADT] Add a new class OwnedOrBorrowed<T>.

FWIW, thanks for the fantastic explanation Zach.

Sep 4 2018, 3:29 AM
chandlerc accepted D51567: CMake: Consolidate gtest detection code.

I mean, sure.

Sep 4 2018, 3:17 AM
chandlerc added a comment to D51317: Handle BumpPtrAllocator::Allocate(0) without undefined behavior.

FWIW, just because a lot of code hits the assert for non-zero size does not (to me) mean that isn't the correct approach.

Sep 4 2018, 3:04 AM

Aug 28 2018

chandlerc added inline comments to D50837: [x86/SLH] Teach SLH to harden against the "ret2spec" attack by implementing the proposed mitigation technique described in the original design document..
Aug 28 2018, 10:33 PM
chandlerc updated the diff for D50837: [x86/SLH] Teach SLH to harden against the "ret2spec" attack by implementing the proposed mitigation technique described in the original design document..

Update addressing review feedback.

Aug 28 2018, 10:33 PM

Aug 27 2018

chandlerc updated the diff for D50837: [x86/SLH] Teach SLH to harden against the "ret2spec" attack by implementing the proposed mitigation technique described in the original design document..

Rebase and ping.

Aug 27 2018, 10:01 AM
chandlerc added a comment to D47858: [New PM] Introducing PassInstrumentation framework.

Very high level comments here...

Aug 27 2018, 1:59 AM
chandlerc committed rL340727: Try to fix this clang driver test case after r340709..
Try to fix this clang driver test case after r340709.
Aug 27 2018, 1:50 AM
chandlerc committed rC340727: Try to fix this clang driver test case after r340709..
Try to fix this clang driver test case after r340709.
Aug 27 2018, 1:50 AM

Aug 26 2018

chandlerc committed rL340721: Fix this file to have the necessary standard library includes and use.
Fix this file to have the necessary standard library includes and use
Aug 26 2018, 11:55 PM
chandlerc committed rL340702: Replace fancy use of initializer lists with simple functions that return.
Replace fancy use of initializer lists with simple functions that return
Aug 26 2018, 3:04 AM
chandlerc committed rL340701: [IR] Replace `isa<TerminatorInst>` with `isTerminator()`..
[IR] Replace `isa<TerminatorInst>` with `isTerminator()`.
Aug 26 2018, 2:52 AM
chandlerc committed rL340700: Avoid specializing a variadic member template in a way that seems to not.
Avoid specializing a variadic member template in a way that seems to not
Aug 26 2018, 2:20 AM
chandlerc committed rL340699: [IR] Sink `isExceptional` predicate to `Instruction`, rename it to.
[IR] Sink `isExceptional` predicate to `Instruction`, rename it to
Aug 26 2018, 1:57 AM
chandlerc committed rL340698: [IR] Begin removal of TerminatorInst by removing successor manipulation..
[IR] Begin removal of TerminatorInst by removing successor manipulation.
Aug 26 2018, 1:43 AM
chandlerc closed D47467: [IR] Begin removal of TerminatorInst by removing successor manipulation..
Aug 26 2018, 1:43 AM
chandlerc added a comment to D47467: [IR] Begin removal of TerminatorInst by removing successor manipulation..

Hi Chandler, are you still planning to work on that?

Thanks, Alexander

Yes, really sorry for the delay here.

This has been held up by finishing off security related work, but should be moving again once the next patch there is finished.

Aug 26 2018, 1:27 AM

Aug 23 2018

chandlerc added inline comments to D51157: [x86/SLH] Add a real Clang flag and LLVM IR attribute for Speculative Load Hardening..
Aug 23 2018, 11:57 PM
chandlerc updated the diff for D51157: [x86/SLH] Add a real Clang flag and LLVM IR attribute for Speculative Load Hardening..

Move to a more conservative model suggested by Kristof.

Aug 23 2018, 11:56 PM
chandlerc committed rC340596: [AVR] Fix inline asm calls now that the addrspace(0) there is explicit..
[AVR] Fix inline asm calls now that the addrspace(0) there is explicit.
Aug 23 2018, 9:46 PM
chandlerc committed rL340596: [AVR] Fix inline asm calls now that the addrspace(0) there is explicit..
[AVR] Fix inline asm calls now that the addrspace(0) there is explicit.
Aug 23 2018, 9:45 PM