Page MenuHomePhabricator

t-tye (Tony Tye)
User

Projects

User does not belong to any projects.

User Details

User Since
Mar 22 2017, 6:01 PM (200 w, 14 h)

Recent Activity

Yesterday

t-tye committed rG079967cdf9e4: [NFC][AMDGPU] Document target ID syntax for code object V2 to V3 (authored by t-tye).
[NFC][AMDGPU] Document target ID syntax for code object V2 to V3
Wed, Jan 20, 1:49 PM
t-tye closed D95018: [NFC][AMDGPU] Document target ID syntax for code object V2 to V3.
Wed, Jan 20, 1:49 PM · Restricted Project
t-tye requested changes to D93525: [OpenMP] Add unbundling of archives containing bundled object files into device specific archives.

At the moment this patch defines compatibility as exact string match of bundler entry ID.
[...]
Supporting target ID requires little more work and discussion.

Let's get this in first, then revisit target ID support as we need it.

Wed, Jan 20, 8:16 AM · Restricted Project

Tue, Jan 19

t-tye requested review of D95018: [NFC][AMDGPU] Document target ID syntax for code object V2 to V3.
Tue, Jan 19, 7:41 PM · Restricted Project

Sun, Jan 17

t-tye added inline comments to D85882: [AMDGPU] Update subtarget features for new target ID support.
Sun, Jan 17, 2:14 PM · Restricted Project
t-tye added inline comments to D85882: [AMDGPU] Update subtarget features for new target ID support.
Sun, Jan 17, 1:05 PM · Restricted Project
t-tye added inline comments to D85882: [AMDGPU] Update subtarget features for new target ID support.
Sun, Jan 17, 12:31 PM · Restricted Project
t-tye added inline comments to D85882: [AMDGPU] Update subtarget features for new target ID support.
Sun, Jan 17, 10:20 AM · Restricted Project

Wed, Jan 13

t-tye added a comment to D93525: [OpenMP] Add unbundling of archives containing bundled object files into device specific archives.

can you document this in ClangOffloadBundler.rst ? I think we need a clear description about how clang-offload-bundler knows which file in the .a file belongs to which target.

How does the .a relate to bundled code objects? Does the .a have a number of bundled code objects? If so wouldn't the identity of code objects be defined by the existing bundled code object ABI already documented? If the .a is a set of non-bundled code objects then defining how they are identified is not part of the clang-offload-bundler documentation as there are no bundled code objects involved. It would seem that the documentation belongs with the OpenMP runtime/compiler that is choosing to use .a files in this manner.

Bundles (created using clang-offload-bundler) are passed to llvm-ar to create an archive of bundled objects (*.a file). An archive can have bundles for multiple device types. So, yes, the identity of code objects is defined by the existing bundled code object ABI.
This patch reads such an archive and produces a device-specific archive for each of the target devices given as input. Each device-specific archive contains all the code objects corresponding to that particular device and are written as per llvm archive format.

Here is a snippet of relevant lit run lines:

// RUN: %clang -O0 -target %itanium_abi_triple %s -c -o %t.o

// RUN: echo 'Content of device file 1' > %t.tgt1
// RUN: clang-offload-bundler -type=o -targets=host-%itanium_abi_triple,openmp-amdgcn-amd-amdhsa-gfx900 -inputs=%t.o,%t.tgt1 -outputs=%t.abundle1.o
 
// RUN: echo 'Content of device file 2' > %t.tgt2
// RUN: clang-offload-bundler -type=o -targets=host-%itanium_abi_triple,openmp-amdgcn-amd-amdhsa-gfx900 -inputs=%t.o,%t.tgt2 -outputs=%t.abundle2.o
 
// RUN: llvm-ar cr %t.lib.a %t.abundle1.o %t.abundle2.o

This patch ==>
// RUN: clang-offload-bundler -unbundle -type=a -targets=openmp-amdgcn-amd-amdhsa-gfx900 -inputs=%t.lib.a -outputs=%t.devicelib.a

%t.devicelib.a will contain all devices objects corresponding to gfx900

Though my interest originates from OpenMP side, Device-specific Archive Libraries created like this can be used by other offloading languages like HIP, CUDA, and OpenCL. Pelase refer D81109 for the an earlier patch in the series of patches which will enable this.

Wed, Jan 13, 8:07 AM · Restricted Project

Tue, Jan 12

t-tye added inline comments to D94528: [libomptarget][amdgpu][nfc] Fix build on centos.
Tue, Jan 12, 11:30 AM · Restricted Project
t-tye added a comment to D93525: [OpenMP] Add unbundling of archives containing bundled object files into device specific archives.

can you document this in ClangOffloadBundler.rst ? I think we need a clear description about how clang-offload-bundler knows which file in the .a file belongs to which target.

Tue, Jan 12, 10:57 AM · Restricted Project

Mon, Jan 11

t-tye committed rG72f1f084670e: [NFC][AMDGPU] Clarify memory model support for volatile (authored by t-tye).
[NFC][AMDGPU] Clarify memory model support for volatile
Mon, Jan 11, 12:00 PM
t-tye closed D94358: [NFC][AMDGPU] Clarify memory model support for volatile.
Mon, Jan 11, 12:00 PM · Restricted Project

Sat, Jan 9

t-tye requested review of D94358: [NFC][AMDGPU] Clarify memory model support for volatile.
Sat, Jan 9, 2:26 AM · Restricted Project

Fri, Jan 8

t-tye committed rG2f499b9affb6: [AMDGPU] Add volatile support to SIMemoryLegalizer (authored by t-tye).
[AMDGPU] Add volatile support to SIMemoryLegalizer
Fri, Jan 8, 4:53 PM
t-tye closed D94214: [AMDGPU] Add volatile support to SIMemoryLegalizer.
Fri, Jan 8, 4:53 PM · Restricted Project
t-tye accepted D94338: [Clang][Docs] Fix ambiguity in clang-offload-bundler docs.

LGTM

Fri, Jan 8, 3:11 PM · Restricted Project

Wed, Jan 6

t-tye requested review of D94214: [AMDGPU] Add volatile support to SIMemoryLegalizer.
Wed, Jan 6, 10:59 PM · Restricted Project

Thu, Dec 24

t-tye committed rG61177943c9c7: [AMDGPU] Use MUBUF instructions for global address space access (authored by pvellien).
[AMDGPU] Use MUBUF instructions for global address space access
Thu, Dec 24, 2:17 AM
t-tye closed D92483: AMDGPU - Use MUBUF instructions for global address space access.
Thu, Dec 24, 2:17 AM · Restricted Project
t-tye accepted D92483: AMDGPU - Use MUBUF instructions for global address space access.
Thu, Dec 24, 12:19 AM · Restricted Project
t-tye accepted D92483: AMDGPU - Use MUBUF instructions for global address space access.

LGTM Thanks.

Thu, Dec 24, 12:10 AM · Restricted Project

Wed, Dec 23

t-tye added inline comments to D92483: AMDGPU - Use MUBUF instructions for global address space access.
Wed, Dec 23, 10:51 PM · Restricted Project
t-tye added inline comments to D92483: AMDGPU - Use MUBUF instructions for global address space access.
Wed, Dec 23, 10:58 AM · Restricted Project
t-tye added inline comments to D92483: AMDGPU - Use MUBUF instructions for global address space access.
Wed, Dec 23, 10:04 AM · Restricted Project
t-tye added inline comments to D92483: AMDGPU - Use MUBUF instructions for global address space access.
Wed, Dec 23, 8:09 AM · Restricted Project

Tue, Dec 22

t-tye added inline comments to D92483: AMDGPU - Use MUBUF instructions for global address space access.
Tue, Dec 22, 1:05 PM · Restricted Project
t-tye added a comment to D92483: AMDGPU - Use MUBUF instructions for global address space access.

I like what Scott describes as it seems intuitively obvious and cleaner. Are there any concerns for making these changes?

Tue, Dec 22, 11:00 AM · Restricted Project

Dec 21 2020

t-tye added a comment to D92483: AMDGPU - Use MUBUF instructions for global address space access.

We did discuss this further and decided that gfx6* should report it supports amdhsa with the restriction that it does not support generic addresses. AMDGPUUsage already notes the restriction in all relevant sections. The LIT tests would also need updating accordingly. The COMgr tests would also need updating.

AMDGPUUsage does not list amdhsa for all 3 SI targets, yet it is going to be accepted with this patch. If that is so then amdhsa has to be listed in the table for these targets.

The table is indicating what runtimes support which ABIs. Since ROCm does not support gfx6* it would still be correct to not list it in the table. It may be useful to have another column to indicate which ABIs are supported by which processors.

Pfff... I do not like it, but OK.

Dec 21 2020, 3:52 PM · Restricted Project
t-tye added a comment to D92483: AMDGPU - Use MUBUF instructions for global address space access.

We did discuss this further and decided that gfx6* should report it supports amdhsa with the restriction that it does not support generic addresses. AMDGPUUsage already notes the restriction in all relevant sections. The LIT tests would also need updating accordingly. The COMgr tests would also need updating.

AMDGPUUsage does not list amdhsa for all 3 SI targets, yet it is going to be accepted with this patch. If that is so then amdhsa has to be listed in the table for these targets.

Dec 21 2020, 2:46 PM · Restricted Project
t-tye added a comment to D92483: AMDGPU - Use MUBUF instructions for global address space access.

But I still do not think this is a right thing to do to accept amdhsa on SI, even if you turn off flat instructions. SI does not support HSA.

As far as I the support of HSA in SI processors. Tony suggested that gfx60x processors should support AMDHSA OS because it is capable of supporting OpenCL 1.2 which not need generic pointers. This topic was discussed in amdgcn weekly and using MUBUF instructions for global address space in gfx60x is decided. The same is also documented in AMDGPU User Guide earlier.

But it is not. In fact AMDGPUUsage.rst says that SI does NOT support amdhsa and it is only supported starting from gfx700:

**GCN GFX6 (Southern Islands (SI))** [AMD-GCN-GFX6]_
-----------------------------------------------------------------------------------------------------------------------
``gfx600``  - ``tahiti``    ``amdgcn``   dGPU                    - Does not      - *pal-amdpal*
                                                                   support
                                                                   generic
                                                                   address
                                                                   space
``gfx601``  - ``pitcairn``  ``amdgcn``   dGPU                    - Does not      - *pal-amdpal*
            - ``verde``                                            support
                                                                   generic
                                                                   address
                                                                   space
``gfx602``  - ``hainan``    ``amdgcn``   dGPU                    - Does not      - *pal-amdpal*
            - ``oland``                                            support
                                                                   generic
                                                                   address
                                                                   space

It is only amdpal listed here.

What was discussed is that we can use buffer instructions for global on SI even if amdhsa is not supported.

At one point, while you were on vacation, the consensus was "support amdhsa OS with gfx6, and fail to compile in the presence of e.g. generic pointers".

Has there been more discussion since then?

Yeah, support amdhsa OS

But I still do not think this is a right thing to do to accept amdhsa on SI, even if you turn off flat instructions. SI does not support HSA.

As far as I the support of HSA in SI processors. Tony suggested that gfx60x processors should support AMDHSA OS because it is capable of supporting OpenCL 1.2 which not need generic pointers. This topic was discussed in amdgcn weekly and using MUBUF instructions for global address space in gfx60x is decided. The same is also documented in AMDGPU User Guide earlier.

But it is not. In fact AMDGPUUsage.rst says that SI does NOT support amdhsa and it is only supported starting from gfx700:

**GCN GFX6 (Southern Islands (SI))** [AMD-GCN-GFX6]_
-----------------------------------------------------------------------------------------------------------------------
``gfx600``  - ``tahiti``    ``amdgcn``   dGPU                    - Does not      - *pal-amdpal*
                                                                   support
                                                                   generic
                                                                   address
                                                                   space
``gfx601``  - ``pitcairn``  ``amdgcn``   dGPU                    - Does not      - *pal-amdpal*
            - ``verde``                                            support
                                                                   generic
                                                                   address
                                                                   space
``gfx602``  - ``hainan``    ``amdgcn``   dGPU                    - Does not      - *pal-amdpal*
            - ``oland``                                            support
                                                                   generic
                                                                   address
                                                                   space

It is only amdpal listed here.

What was discussed is that we can use buffer instructions for global on SI even if amdhsa is not supported.

At one point, while you were on vacation, the consensus was "support amdhsa OS with gfx6, and fail to compile in the presence of e.g. generic pointers".

Has there been more discussion since then?

I see that is not what we document. @t-tye whatis your opinion?

@scott.linder there is no discussion since then.
@rampitec the final consensus is to use buffer instructions for global address space accesses in SI in presence of amdhsa. In other OS types for SI, backend generates buffer instructions for global addr space access already.

Dec 21 2020, 2:35 PM · Restricted Project
t-tye accepted D93648: Revert "[AMDGPU][HIP] Switch default DWARF version to 5".

LGTM

Dec 21 2020, 1:41 PM · Restricted Project

Dec 17 2020

t-tye accepted D93125: Update AMDGPU PAL usage documentation.

LGTM

Dec 17 2020, 1:11 PM · Restricted Project
t-tye edited reviewers for D93125: Update AMDGPU PAL usage documentation, added: t-tye; removed: tony-tye.
Dec 17 2020, 12:14 PM · Restricted Project
t-tye added inline comments to D93125: Update AMDGPU PAL usage documentation.
Dec 17 2020, 12:13 PM · Restricted Project
t-tye committed rG8c6d516286d5: [NFC][AMDGPU] Reorganize description of scratch handling (authored by t-tye).
[NFC][AMDGPU] Reorganize description of scratch handling
Dec 17 2020, 11:34 AM
t-tye closed D93440: [NFC][AMDGPU] Format change to processr table in AMGPUUsage.rst.
Dec 17 2020, 11:34 AM · Restricted Project
t-tye updated the diff for D93440: [NFC][AMDGPU] Format change to processr table in AMGPUUsage.rst.

Reorganize scratch description.

Dec 17 2020, 12:49 AM · Restricted Project

Dec 16 2020

t-tye updated the diff for D93440: [NFC][AMDGPU] Format change to processr table in AMGPUUsage.rst.

Correct typo.

Dec 16 2020, 7:06 PM · Restricted Project
t-tye requested review of D93440: [NFC][AMDGPU] Format change to processr table in AMGPUUsage.rst.
Dec 16 2020, 6:49 PM · Restricted Project
t-tye accepted D93398: [NFC] Use regex for code object version in hip tests.

LGTM

Dec 16 2020, 8:58 AM · Restricted Project
t-tye added inline comments to D93398: [NFC] Use regex for code object version in hip tests.
Dec 16 2020, 7:58 AM · Restricted Project
t-tye added a reviewer for D93398: [NFC] Use regex for code object version in hip tests: t-tye.
Dec 16 2020, 7:57 AM · Restricted Project
t-tye added inline comments to D93398: [NFC] Use regex for code object version in hip tests.
Dec 16 2020, 7:57 AM · Restricted Project

Dec 15 2020

t-tye accepted D93353: [AMDGPU] Print SCRATCH_EN field after the kernel.

LGTM

Dec 15 2020, 6:31 PM · Restricted Project
t-tye committed rGd5ea8f70105a: [AMDGPU] Clarify scratch initialization (authored by t-tye).
[AMDGPU] Clarify scratch initialization
Dec 15 2020, 12:15 PM
t-tye closed D93271: [AMDGPU] Clarify scratch initialization.
Dec 15 2020, 12:15 PM · Restricted Project

Dec 14 2020

t-tye updated the diff for D93271: [AMDGPU] Clarify scratch initialization.

Fix missing blank line.

Dec 14 2020, 10:42 PM · Restricted Project
t-tye requested review of D93271: [AMDGPU] Clarify scratch initialization.
Dec 14 2020, 8:53 PM · Restricted Project
t-tye accepted D93258: [amdgpu] Default to code object v3.

LGTM

Dec 14 2020, 5:09 PM · Restricted Project, Restricted Project
t-tye added inline comments to D93258: [amdgpu] Default to code object v3.
Dec 14 2020, 5:06 PM · Restricted Project, Restricted Project
t-tye added reviewers for D93258: [amdgpu] Default to code object v3: yaxunl, kzhuravl.
Dec 14 2020, 4:16 PM · Restricted Project, Restricted Project
t-tye committed rG90b951dd6801: [NFC] Remove trailing whitespace in llvm/CMakeLists.txt (authored by t-tye).
[NFC] Remove trailing whitespace in llvm/CMakeLists.txt
Dec 14 2020, 9:51 AM
t-tye closed D93234: [NFC] Remove trailing whitespace in llvm/CMakeLists.txt.
Dec 14 2020, 9:51 AM · Restricted Project
t-tye requested review of D93234: [NFC] Remove trailing whitespace in llvm/CMakeLists.txt.
Dec 14 2020, 9:48 AM · Restricted Project

Dec 13 2020

t-tye committed rG5ad202ce8963: [NFC][AMDGPU] Reformat AMD GPU targets in cuda.cpp (authored by t-tye).
[NFC][AMDGPU] Reformat AMD GPU targets in cuda.cpp
Dec 13 2020, 3:04 PM
t-tye closed D93181: [NFC][AMDGPU] Reformat AMD GPU targets in cuda.cpp.
Dec 13 2020, 3:03 PM · Restricted Project
t-tye requested review of D93181: [NFC][AMDGPU] Reformat AMD GPU targets in cuda.cpp.
Dec 13 2020, 9:26 AM · Restricted Project
t-tye committed rG828602c772c8: [NFC]{AMDGPU] Update AMDGPUUsage with AMD RDNA 2 reference (authored by t-tye).
[NFC]{AMDGPU] Update AMDGPUUsage with AMD RDNA 2 reference
Dec 13 2020, 9:22 AM
t-tye closed D93172: [NFC]{AMDGPU] Update AMDGPUUsage with AMD RDNA 2 reference.
Dec 13 2020, 9:22 AM · Restricted Project

Dec 12 2020

t-tye updated the summary of D93172: [NFC]{AMDGPU] Update AMDGPUUsage with AMD RDNA 2 reference.
Dec 12 2020, 5:52 PM · Restricted Project
t-tye updated the diff for D93172: [NFC]{AMDGPU] Update AMDGPUUsage with AMD RDNA 2 reference.

Update commit message.

Dec 12 2020, 5:52 PM · Restricted Project
t-tye requested review of D93172: [NFC]{AMDGPU] Update AMDGPUUsage with AMD RDNA 2 reference.
Dec 12 2020, 5:22 PM · Restricted Project
t-tye committed rG7beee561e23d: [AMDGPU] Add missing targets to target-invalid-cpu-note.c (authored by t-tye).
[AMDGPU] Add missing targets to target-invalid-cpu-note.c
Dec 12 2020, 10:41 AM
t-tye committed rG92ab6ed6672b: [AMDGPU] Add missing targets to amdgpu-features.cl (authored by t-tye).
[AMDGPU] Add missing targets to amdgpu-features.cl
Dec 12 2020, 10:41 AM
t-tye committed rG87a4e14e40ee: [NFC][AMDGPU] AMDGPUUsage updates (authored by t-tye).
[NFC][AMDGPU] AMDGPUUsage updates
Dec 12 2020, 10:41 AM
t-tye closed D93018: [AMDGPU] Add missing targets to target-invalid-cpu-note.c.
Dec 12 2020, 10:41 AM · Restricted Project
t-tye closed D93017: [AMDGPU] Add missing targets to amdgpu-features.cl.
Dec 12 2020, 10:41 AM · Restricted Project
t-tye closed D93016: [NFC][AMDGPU] AMDGPUUsage updates.
Dec 12 2020, 10:41 AM · Restricted Project

Dec 10 2020

t-tye requested review of D93018: [AMDGPU] Add missing targets to target-invalid-cpu-note.c.
Dec 10 2020, 2:44 AM · Restricted Project
t-tye requested review of D93017: [AMDGPU] Add missing targets to amdgpu-features.cl.
Dec 10 2020, 2:43 AM · Restricted Project
t-tye requested review of D93016: [NFC][AMDGPU] AMDGPUUsage updates.
Dec 10 2020, 2:42 AM · Restricted Project

Dec 4 2020

t-tye committed rG3242eaef2719: [NFC][AMDGPU] AMDGPUUsage updates (authored by t-tye).
[NFC][AMDGPU] AMDGPUUsage updates
Dec 4 2020, 6:14 PM
t-tye closed D92708: [NFC][AMDGPU] AMDGPUUsage updates.
Dec 4 2020, 6:14 PM · Restricted Project
t-tye requested review of D92708: [NFC][AMDGPU] AMDGPUUsage updates.
Dec 4 2020, 4:59 PM · Restricted Project

Dec 2 2020

t-tye requested changes to D92483: AMDGPU - Use MUBUF instructions for global address space access.
Dec 2 2020, 11:36 AM · Restricted Project

Dec 1 2020

t-tye committed rGac1b2ae9dc6c: [NFC][AMDGPU] Fix broken link to ClangOffloadBundler in AMDGPUUsage (authored by t-tye).
[NFC][AMDGPU] Fix broken link to ClangOffloadBundler in AMDGPUUsage
Dec 1 2020, 7:06 PM
t-tye committed rGa417cb086209: [NFC] Add CLangOffloadBundler documentation to Clang index (authored by t-tye).
[NFC] Add CLangOffloadBundler documentation to Clang index
Dec 1 2020, 6:48 PM
t-tye closed D92441: Add CLangOffloadBundler documentation to Clang index.
Dec 1 2020, 6:48 PM · Restricted Project
t-tye requested review of D92441: Add CLangOffloadBundler documentation to Clang index.
Dec 1 2020, 6:08 PM · Restricted Project
t-tye committed rG04424c69bc7b: [NFC][AMDGPU] AMDGPU code object V4 ABI documentation (authored by t-tye).
[NFC][AMDGPU] AMDGPU code object V4 ABI documentation
Dec 1 2020, 3:39 PM
t-tye closed D92434: [NFC][AMDGPU] AMDGPU code object V4 ABI documentation.
Dec 1 2020, 3:39 PM · Restricted Project, Restricted Project
t-tye requested review of D92434: [NFC][AMDGPU] AMDGPU code object V4 ABI documentation.
Dec 1 2020, 3:31 PM · Restricted Project, Restricted Project

Nov 30 2020

t-tye added inline comments to D85882: [AMDGPU] Update subtarget features for new target ID support.
Nov 30 2020, 1:59 PM · Restricted Project
t-tye added inline comments to D85882: [AMDGPU] Update subtarget features for new target ID support.
Nov 30 2020, 1:39 PM · Restricted Project

Nov 27 2020

t-tye added inline comments to D92115: AMDGPU - Add diagnostic for compiling modules with AMD HSA OS type and GFX 6 arch .
Nov 27 2020, 11:36 AM · Restricted Project, Restricted Project
t-tye requested changes to D92115: AMDGPU - Add diagnostic for compiling modules with AMD HSA OS type and GFX 6 arch .
Nov 27 2020, 9:43 AM · Restricted Project, Restricted Project

Nov 20 2020

t-tye committed rG8605d3134c97: [NFC][AMDGPU] Document kernel descriptor (authored by t-tye).
[NFC][AMDGPU] Document kernel descriptor
Nov 20 2020, 8:55 PM
t-tye closed D91458: [NFC][AMDGPU] Document kernel descriptor.
Nov 20 2020, 8:55 PM · Restricted Project
t-tye accepted D91545: [AMDGPU] Extend and reorganize memory legalizer tests.

LGTM

Nov 20 2020, 3:00 PM · Restricted Project
t-tye accepted D91545: [AMDGPU] Extend and reorganize memory legalizer tests.

Reviewed script that produces the new organization of the tests.

Nov 20 2020, 1:23 PM · Restricted Project

Nov 13 2020

t-tye requested review of D91458: [NFC][AMDGPU] Document kernel descriptor.
Nov 13 2020, 1:13 PM · Restricted Project

Nov 9 2020

t-tye added inline comments to D91086: AMDGPU: Document why we use (non-volatile) BUFFER_WBINVL1 in graphics.
Nov 9 2020, 11:24 AM · Restricted Project

Nov 3 2020

t-tye committed rG45bcbe46d75a: [NFC][AMDGPU] Minor editorial improvements to AMDGPUUsage.rst (authored by t-tye).
[NFC][AMDGPU] Minor editorial improvements to AMDGPUUsage.rst
Nov 3 2020, 8:57 AM
t-tye closed D90661: [NFC][AMDGPU] Minor editorial improvements to AMDGPUUsage.rst.
Nov 3 2020, 8:56 AM · Restricted Project

Nov 2 2020

t-tye requested review of D90661: [NFC][AMDGPU] Minor editorial improvements to AMDGPUUsage.rst.
Nov 2 2020, 6:28 PM · Restricted Project
t-tye added inline comments to D90548: [NFC][AMDGPU] Restructure the AMDGPU memory model description.
Nov 2 2020, 6:27 PM · Restricted Project
t-tye committed rG68160789c120: [NFC][AMDGPU] Restructure the AMDGPU memory model description (authored by t-tye).
[NFC][AMDGPU] Restructure the AMDGPU memory model description
Nov 2 2020, 1:33 PM
t-tye closed D90548: [NFC][AMDGPU] Restructure the AMDGPU memory model description.
Nov 2 2020, 1:33 PM · Restricted Project
t-tye updated the diff for D90548: [NFC][AMDGPU] Restructure the AMDGPU memory model description.

Correct typos.

Nov 2 2020, 1:23 PM · Restricted Project