Page MenuHomePhabricator
People lei

lei (Lei Huang)
User

Projects

User does not belong to any projects.

User Details

User Since
Oct 24 2016, 11:46 AM (151 w, 3 d)

Recent Activity
View All

Mon, Sep 16

lei committed rGbfb197d7a3b3: [PowerPC] Cust lower fpext v2f32 to v2f64 from extract_subvector v4f32 (authored by lei).
[PowerPC] Cust lower fpext v2f32 to v2f64 from extract_subvector v4f32
Mon, Sep 16, 1:05 PM
lei committed rL372029: [PowerPC] Cust lower fpext v2f32 to v2f64 from extract_subvector v4f32.
[PowerPC] Cust lower fpext v2f32 to v2f64 from extract_subvector v4f32
Mon, Sep 16, 1:05 PM
lei closed D61961: [PowerPC] Cust lower fpext v2f32 to v2f64 from extract_subvector v4f32.
Mon, Sep 16, 1:05 PM · Restricted Project

Thu, Sep 12

lei added inline comments to D60506: [CGP] Make ICMP_EQ use CR result of ICMP_S(L|G)T dominators.
Thu, Sep 12, 4:45 PM · Restricted Project
lei added a comment to D61961: [PowerPC] Cust lower fpext v2f32 to v2f64 from extract_subvector v4f32.

@nemanjai All comments have been addressed. Please take another look. Thx!

Thu, Sep 12, 12:36 PM · Restricted Project

Wed, Sep 11

lei added inline comments to D63676: Disable hosting MI to hotter basic blocks.
Wed, Sep 11, 7:28 AM · Restricted Project
lei updated the diff for D61961: [PowerPC] Cust lower fpext v2f32 to v2f64 from extract_subvector v4f32.

Addressed reviewer comments.

Wed, Sep 11, 7:02 AM · Restricted Project

Jul 23 2019

lei committed rG686cee09450e: [NFC][ASAN] Add brackets around not command (authored by lei).
[NFC][ASAN] Add brackets around not command
Jul 23 2019, 6:12 AM
lei committed rL366805: [NFC][ASAN] Add brackets around not command.
[NFC][ASAN] Add brackets around not command
Jul 23 2019, 6:12 AM

Jul 22 2019

lei updated the diff for D61961: [PowerPC] Cust lower fpext v2f32 to v2f64 from extract_subvector v4f32.

Address review comments.

Jul 22 2019, 11:49 AM · Restricted Project

Jul 5 2019

lei added a comment to D61961: [PowerPC] Cust lower fpext v2f32 to v2f64 from extract_subvector v4f32.

Ping

Jul 5 2019, 9:40 AM · Restricted Project

Jun 12 2019

lei added inline comments to D61249: [NFC] Refactor the systemZ scheduler strategy to adapt with the new schedule state forwarding infrastructure.
Jun 12 2019, 7:43 AM · Restricted Project
lei added inline comments to D61248: [NFC] Add the infrastructure to forward the scheduled state between MBB.
Jun 12 2019, 5:17 AM · Restricted Project

May 15 2019

lei updated the diff for D61961: [PowerPC] Cust lower fpext v2f32 to v2f64 from extract_subvector v4f32.
May 15 2019, 1:46 PM · Restricted Project
lei created D61961: [PowerPC] Cust lower fpext v2f32 to v2f64 from extract_subvector v4f32.
May 15 2019, 1:37 PM · Restricted Project

May 14 2019

lei committed rG22561972af47: [PowerPC] Custom lower known CR bit spills (authored by lei).
[PowerPC] Custom lower known CR bit spills
May 14 2019, 7:25 AM
lei committed rL360677: [PowerPC] Custom lower known CR bit spills.
[PowerPC] Custom lower known CR bit spills
May 14 2019, 7:24 AM
lei closed D61754: [PowerPC] Custom lower known CR bit spills.
May 14 2019, 7:24 AM · Restricted Project
lei added a comment to D61754: [PowerPC] Custom lower known CR bit spills.
In D61754#1500658, @hfinkel wrote:

LGTM.

If this test case proves too fragile in the future, we might replace it with an MIR test.

May 14 2019, 6:45 AM · Restricted Project

May 13 2019

lei added a comment to D61754: [PowerPC] Custom lower known CR bit spills.

address spacing issues.

May 13 2019, 1:11 PM · Restricted Project
lei updated the diff for D61754: [PowerPC] Custom lower known CR bit spills.
May 13 2019, 1:11 PM · Restricted Project
lei updated the diff for D61754: [PowerPC] Custom lower known CR bit spills.

Added option to specify cutoff for CR bit definition search with default of 100.

May 13 2019, 12:45 PM · Restricted Project

May 10 2019

lei added inline comments to D61754: [PowerPC] Custom lower known CR bit spills.
May 10 2019, 3:35 PM · Restricted Project
lei added inline comments to D61754: [PowerPC] Custom lower known CR bit spills.
May 10 2019, 12:47 PM · Restricted Project
lei committed rG1ac6e9636c9e: [PowerPC] custom lower `v2f64 fpext v2f32` (authored by lei).
[PowerPC] custom lower `v2f64 fpext v2f32`
May 10 2019, 7:03 AM
lei committed rL360429: [PowerPC] custom lower `v2f64 fpext v2f32`.
[PowerPC] custom lower `v2f64 fpext v2f32`
May 10 2019, 7:02 AM
lei closed D57857: [PowerPC] custom lower `v2f64 fpext v2f32`.
May 10 2019, 7:02 AM · Restricted Project

May 9 2019

lei created D61754: [PowerPC] Custom lower known CR bit spills.
May 9 2019, 12:38 PM · Restricted Project

Apr 8 2019

lei requested review of D57857: [PowerPC] custom lower `v2f64 fpext v2f32`.
Apr 8 2019, 11:56 AM · Restricted Project
lei updated the diff for D57857: [PowerPC] custom lower `v2f64 fpext v2f32`.

Update comments and renamed new ppc ISD nodes.

Apr 8 2019, 11:41 AM · Restricted Project

Feb 13 2019

lei updated the diff for D57857: [PowerPC] custom lower `v2f64 fpext v2f32`.

Address review comments.

Feb 13 2019, 11:42 AM · Restricted Project
lei requested review of D57857: [PowerPC] custom lower `v2f64 fpext v2f32`.
Feb 13 2019, 5:21 AM · Restricted Project

Feb 6 2019

lei created D57857: [PowerPC] custom lower `v2f64 fpext v2f32`.
Feb 6 2019, 3:46 PM · Restricted Project
lei added a comment to D57535: [CMake] Use LLVM_ENABLE_PROJECTS as the "single source" of truth when used..

@delcypher I am having a problem with my internal build after this change have gone in. Can you please help to determine what went wrong? Previously I was able to do the following:

$ cmake -G Ninja -DCMAKE_BUILD_TYPE=Release -DLLVM_ENABLE_ASSERTIONS=On '-DLLVM_ENABLE_PROJECTS=clang;llvm;compiler-rt;clang-tools-extra;openmp;lld' -DCMAKE_C_COMPILER=/home/llvm/gcc/7.3.0/bin/gcc -DCMAKE_CXX_COMPILER=/home/llvm/gcc/7.3.0/bin/g++ -DLLVM_BINUTILS_INCDIR=/usr/include $LLVM_SRC
$ ninja -j 20
$ ninja check-libomptarget check-openmp
Feb 6 2019, 1:02 PM · Restricted Project

Nov 2 2018

lei committed rCRT346030: [PowerPC]Disable randomized address space on Linux ppc64le.
[PowerPC]Disable randomized address space on Linux ppc64le
Nov 2 2018, 12:03 PM
lei committed rL346030: [PowerPC]Disable randomized address space on Linux ppc64le.
[PowerPC]Disable randomized address space on Linux ppc64le
Nov 2 2018, 12:02 PM
lei closed D52900: [PowerPC]Disable randomized address space on Linux ppc64le.
Nov 2 2018, 12:02 PM
lei added a reviewer for D52900: [PowerPC]Disable randomized address space on Linux ppc64le: nemanjai.
Nov 2 2018, 7:50 AM

Oct 26 2018

lei committed rL345414: [PowerPC] Improve BUILD_VECTOR of 4 i32s.
[PowerPC] Improve BUILD_VECTOR of 4 i32s
Oct 26 2018, 11:13 AM
lei closed D53494: [PowerPC] Improve BUILD_VECTOR of 4 i32s.
Oct 26 2018, 11:13 AM

Oct 22 2018

lei created D53494: [PowerPC] Improve BUILD_VECTOR of 4 i32s.
Oct 22 2018, 6:15 AM

Oct 17 2018

lei added a comment to D22435: Fix pruned jump table basic blocks causing compilation failure (Bug 28102).

PR39124 has a reduced IR that seem to be the same issue this fixes. However I think Chri's suggested fix in PR28102 to remove the MBB references from the jump tables is a better way to go. Can you please give that a try?

Oct 17 2018, 6:11 AM

Oct 15 2018

lei added a comment to D52900: [PowerPC]Disable randomized address space on Linux ppc64le.

Any other concerns about committing this patch?

Oct 15 2018, 6:50 AM

Oct 9 2018

lei updated the diff for D52900: [PowerPC]Disable randomized address space on Linux ppc64le.

Warning message updated.

Oct 9 2018, 11:59 AM
lei added a comment to D52900: [PowerPC]Disable randomized address space on Linux ppc64le.

For PPC this needs to be done for both tsan and msan.

Oct 9 2018, 7:22 AM

Oct 4 2018

lei created D52900: [PowerPC]Disable randomized address space on Linux ppc64le.
Oct 4 2018, 12:01 PM

Sep 25 2018

lei added a comment to rL343015: [Sanitizers] [MinGW] Produce undecorated symbols for /export: directives when….

This is causing the buildbot failures for powerpc: http://lab.llvm.org:8011/builders/clang-ppc64le-linux

Sep 25 2018, 2:25 PM

Sep 24 2018

lei added a comment to rL342657: [gcov] Fix wrong line hit counts when multiple blocks are on the same line.
In rL342657#289221, @calixte wrote:

According to the logs, there is an issue in the # of Runs for each file which should be fixed in clearing directory build/projects/compiler-rt/test/profile/Profile-x86_64/Output/ on build machines.
@lei how can I do that ?

Sep 24 2018, 9:38 AM
lei added a comment to rL342657: [gcov] Fix wrong line hit counts when multiple blocks are on the same line.

This has broken ALL of the following build bots and they have been broken for more than 3 days. Please fix this or pull the changeset:

Sep 24 2018, 7:20 AM

Aug 13 2018

lei accepted D49080: [PowerPC] Don't run BV DAG Combine before legalization if it assumes legal types.

LGTM

Aug 13 2018, 10:13 AM

Aug 10 2018

lei added inline comments to D49507: [Power9] Add __float128 support in the backend for bitcast to a i128.
Aug 10 2018, 6:15 AM

Jul 5 2018

lei committed rL336361: [Power9] Add lib calls for float128 operations with no equivalent PPC….
[Power9] Add lib calls for float128 operations with no equivalent PPC…
Jul 5 2018, 8:26 AM
lei closed D48544: Add lib calls for float128 operations with no equivalent PPC instruction.
Jul 5 2018, 8:26 AM
lei committed rL336316: [Power9] Optimize codgen for conversions of int to float128.
[Power9] Optimize codgen for conversions of int to float128
Jul 5 2018, 12:51 AM
lei closed D48429: [Power9] Optimize codgen for conversions of int to float128.
Jul 5 2018, 12:50 AM

Jul 4 2018

lei committed rL336314: [Power9][NFC] add back-end tests for passing homogeneous fp128 aggregates by….
[Power9][NFC] add back-end tests for passing homogeneous fp128 aggregates by…
Jul 4 2018, 11:56 PM
lei closed D48310: [Power9][NFC] add back-end test case for passing homogeneous fp128 aggregates by value.
Jul 4 2018, 11:56 PM
lei committed rL336313: [Power9] Add tests for passing float128 in VSX reg for non-homogenous aggregates.
[Power9] Add tests for passing float128 in VSX reg for non-homogenous aggregates
Jul 4 2018, 11:34 PM
lei committed rL336310: [Power9] Ensure float128 in non-homogenous aggregates are passed via VSX reg.
[Power9] Ensure float128 in non-homogenous aggregates are passed via VSX reg
Jul 4 2018, 11:26 PM
lei closed D48308: [Power9] Ensure float128 in non-homogenous aggregates are passed via VSX registers.
Jul 4 2018, 11:26 PM
lei added inline comments to D48308: [Power9] Ensure float128 in non-homogenous aggregates are passed via VSX registers.
Jul 4 2018, 10:54 PM
lei committed rC336308: [Power9] Update fp128 as a valid homogenous aggregate base type.
[Power9] Update fp128 as a valid homogenous aggregate base type
Jul 4 2018, 9:37 PM
lei committed rL336308: [Power9] Update fp128 as a valid homogenous aggregate base type.
[Power9] Update fp128 as a valid homogenous aggregate base type
Jul 4 2018, 9:36 PM
lei closed D48044: [Power9] Update fp128 as a valid homogenous aggregate base type .
Jul 4 2018, 9:36 PM
lei committed rL336307: [Power9]Legalize and emit code for quad-precision convert from single-precision.
[Power9]Legalize and emit code for quad-precision convert from single-precision
Jul 4 2018, 9:23 PM
lei closed D47569: [Power9]Legalize and emit code for quad-precision convert from single-precision.
Jul 4 2018, 9:23 PM
lei committed rL336306: [Power9] Implement float128 parameter passing and return values.
[Power9] Implement float128 parameter passing and return values
Jul 4 2018, 9:15 PM
lei closed D47552: [Power9] Implement float128 parameter passing and return values.
Jul 4 2018, 9:15 PM
lei committed rL336299: [Power9]Legalize and emit code for round & convert quad-precision values.
[Power9]Legalize and emit code for round & convert quad-precision values
Jul 4 2018, 3:04 PM
lei closed D46997: [Power9]Legalize and emit code for round & convert quad-precision values.
Jul 4 2018, 3:04 PM

Jun 28 2018

lei abandoned D47573: [Power9] Enable extload of constant values to f128.

merged into D47552

Jun 28 2018, 1:03 PM

Jun 26 2018

lei added a comment to rL335514: [PowerPC] Fix incorrectly encoded wait instruction.

I am not sure why this has changed since ISA 2.07. But I have verified the new encoding on pwr8/pwr9 and that the old encoding produces an illegal instruction.

Jun 26 2018, 1:24 PM
lei accepted D48483: [Power9] [LLVM] Add __float128 support for trunc to double round to odd.

LGTM. Please feel free to fix the nit comments before committing.

Jun 26 2018, 11:33 AM
lei accepted D48482: [Power9] [CLANG] Add __float128 support for trunc to double round to odd.
Jun 26 2018, 11:26 AM
lei added a comment to D48482: [Power9] [CLANG] Add __float128 support for trunc to double round to odd.

LGTM

Jun 26 2018, 11:25 AM
lei added inline comments to D48185: [Power9] [LLVM] Add __float128 exponent GET and SET builtins.
Jun 26 2018, 9:04 AM
lei added inline comments to D48218: [Power9] Add remaining __flaot128 builtin support for FMA round to odd.
Jun 26 2018, 6:36 AM

Jun 25 2018

lei committed rL335514: [PowerPC] Fix incorrectly encoded wait instruction.
[PowerPC] Fix incorrectly encoded wait instruction
Jun 25 2018, 12:33 PM
lei closed D48550: [PowerPC] Fix incorrectly encoded wait instruction.
Jun 25 2018, 12:33 PM
lei created D48550: [PowerPC] Fix incorrectly encoded wait instruction.
Jun 25 2018, 9:04 AM
lei added a parent revision for D48544: Add lib calls for float128 operations with no equivalent PPC instruction: D47552: [Power9] Implement float128 parameter passing and return values.
Jun 25 2018, 6:02 AM
lei added a child revision for D47552: [Power9] Implement float128 parameter passing and return values: D48544: Add lib calls for float128 operations with no equivalent PPC instruction.
Jun 25 2018, 6:02 AM
lei created D48544: Add lib calls for float128 operations with no equivalent PPC instruction.
Jun 25 2018, 6:01 AM

Jun 21 2018

lei created D48429: [Power9] Optimize codgen for conversions of int to float128.
Jun 21 2018, 6:55 AM

Jun 19 2018

lei added inline comments to D48184: [Power9] [CLANG] Add __float128 exponent GET and SET builtins.
Jun 19 2018, 12:13 PM
lei added a child revision for D47552: [Power9] Implement float128 parameter passing and return values: D47569: [Power9]Legalize and emit code for quad-precision convert from single-precision.
Jun 19 2018, 7:34 AM
lei added a parent revision for D47569: [Power9]Legalize and emit code for quad-precision convert from single-precision: D47552: [Power9] Implement float128 parameter passing and return values.
Jun 19 2018, 7:34 AM
lei added a parent revision for D47573: [Power9] Enable extload of constant values to f128: D47552: [Power9] Implement float128 parameter passing and return values.
Jun 19 2018, 7:33 AM
lei added a parent revision for D48308: [Power9] Ensure float128 in non-homogenous aggregates are passed via VSX registers: D47552: [Power9] Implement float128 parameter passing and return values.
Jun 19 2018, 7:33 AM
lei added child revisions for D47552: [Power9] Implement float128 parameter passing and return values: D47573: [Power9] Enable extload of constant values to f128, D48044: [Power9] Update fp128 as a valid homogenous aggregate base type , D48308: [Power9] Ensure float128 in non-homogenous aggregates are passed via VSX registers.
Jun 19 2018, 7:33 AM
lei added a parent revision for D48044: [Power9] Update fp128 as a valid homogenous aggregate base type : D47552: [Power9] Implement float128 parameter passing and return values.
Jun 19 2018, 7:33 AM
lei added a parent revision for D48310: [Power9][NFC] add back-end test case for passing homogeneous fp128 aggregates by value: D48044: [Power9] Update fp128 as a valid homogenous aggregate base type .
Jun 19 2018, 7:30 AM
lei added a child revision for D48044: [Power9] Update fp128 as a valid homogenous aggregate base type : D48310: [Power9][NFC] add back-end test case for passing homogeneous fp128 aggregates by value.
Jun 19 2018, 7:30 AM
lei retitled D48308: [Power9] Ensure float128 in non-homogenous aggregates are passed via VSX registers from [POWER9] Ensure float128 in non-homogenous aggregates are passed via VSX registers to [Power9] Ensure float128 in non-homogenous aggregates are passed via VSX registers.
Jun 19 2018, 6:08 AM
lei retitled D48310: [Power9][NFC] add back-end test case for passing homogeneous fp128 aggregates by value from [POWER9][NFC] add back-end test case for passing homogeneous fp128 aggregates by value to [Power9][NFC] add back-end test case for passing homogeneous fp128 aggregates by value.
Jun 19 2018, 6:07 AM

Jun 18 2018

lei added a comment to D47550: [Power9] Add __float128 builtins for Round To Odd.

The indentation in the td files are still wrong. You should take a look at the existing pattern indentations and follow ...

Jun 18 2018, 9:01 PM
lei created D48310: [Power9][NFC] add back-end test case for passing homogeneous fp128 aggregates by value.
Jun 18 2018, 8:40 PM
lei created D48308: [Power9] Ensure float128 in non-homogenous aggregates are passed via VSX registers.
Jun 18 2018, 8:08 PM
lei added inline comments to D48294: [Power9] Add __float128 support for compare operations.
Jun 18 2018, 1:25 PM
lei added inline comments to D48294: [Power9] Add __float128 support for compare operations.
Jun 18 2018, 1:23 PM