[qt/clang/llvm-project]: Summary of bulk changes made

Git repository change summary for qt/clang/llvm-project
Pushed by mirror-service into branch 'upstream/users/cdevadas/subreg-reload'.
Changed from 070b3e99ac751c9e6cb38201e0eec68d72a55542 to 829c8e9360ece47f2ec76c67ee0f78306e2312cd
Acknowledgement was received that this change introduces only existing code that has been pushed to another public open source repository.

This change contains the following new commits:

Git commit 19a7d30a4936a05af0e92be7082db0106066bc84 by Aaditya (on behalf of Christudasan Devadasan) on 30/07/2026 at 06:10..
[AMDGPU] Introduce Offset field in SGPR spill Pseudos

Currently, SGPR spill pseudo-instructions lack
an offset field to represent non-zero stack offsets.
This patch introduces an additional offset field to
SGPR spill pseudo-instructions and updates all
relevant passes that handle spill lowering to support
this new field. This field is essential for a future
patch that implements subreg reload of tuple registers
from their stack location during RA.
https://invent.kde.org/qt/clang/llvm-project/-/commit/19a7d30a4936a05af0e92be7082db0106066bc84

Git commit e656f3036ad6a4793bdc27bb855636dc01538ba9 by Aaditya (on behalf of Christudasan Devadasan) on 30/07/2026 at 06:10..
incorporated review comments.
https://invent.kde.org/qt/clang/llvm-project/-/commit/e656f3036ad6a4793bdc27bb855636dc01538ba9

Git commit 76d90cd95bbd9b968a5a4f6e1190e58455fc7006 by Aaditya (on behalf of Christudasan Devadasan) on 30/07/2026 at 06:20..
[AMDGPU] Make getNumSubRegsForSpillOp externally available (NFC).
https://invent.kde.org/qt/clang/llvm-project/-/commit/76d90cd95bbd9b968a5a4f6e1190e58455fc7006

Git commit bc1cf7619e13569da0809968f52963d2a9f96810 by Aaditya (on behalf of Christudasan Devadasan) on 30/07/2026 at 06:26..
moved the implementation to SIInstrInfo.
https://invent.kde.org/qt/clang/llvm-project/-/commit/bc1cf7619e13569da0809968f52963d2a9f96810

Git commit 836607c70fa77dcc960e30d4ecd7ced2c67ff084 by Aaditya (on behalf of Christudasan Devadasan) on 30/07/2026 at 06:26..
fixed a comment.
https://invent.kde.org/qt/clang/llvm-project/-/commit/836607c70fa77dcc960e30d4ecd7ced2c67ff084

Git commit c2eb408cf1e7ddab8a92c4f5310a233210d06df0 by Aaditya (on behalf of Christudasan Devadasan) on 30/07/2026 at 06:26..
[AMDGPU] Make AMDGPURewriteAGPRCopyMFMA aware of subreg reload

AMDGPURewriteAGPRCopyMFMA pass is currently not subreg-aware.
In particular, the logic that optimizes spills into COPY
instructions assumes full register reloads. This becomes
problematic when the reload instruction partially restores
a tuple register. This patch introduces the necessary changes
to make this pass subreg-aware, for a future patch that
implements subreg reload during RA.
https://invent.kde.org/qt/clang/llvm-project/-/commit/c2eb408cf1e7ddab8a92c4f5310a233210d06df0

Git commit 5227fe6e892f4f108401604fbe563fe32055914c by Aaditya (on behalf of Christudasan Devadasan) on 30/07/2026 at 06:26..
suggestions incorporated.
https://invent.kde.org/qt/clang/llvm-project/-/commit/5227fe6e892f4f108401604fbe563fe32055914c

Git commit 2a5c8c9adfb79219e6a07bf6780c2c7daa5e8a9b by Aaditya on 31/07/2026 at 08:57..
Fix Merge conflict from rebasing.
https://invent.kde.org/qt/clang/llvm-project/-/commit/2a5c8c9adfb79219e6a07bf6780c2c7daa5e8a9b

Git commit bb5f7a2a89e2662aaaef38717ab3a0352e77518a by Aaditya (on behalf of Christudasan Devadasan) on 31/07/2026 at 08:57..
[CodeGen] Enhance createFrom for sub-reg aware cloning

Instead of just cloning the virtual register, this
function now creates a new virtual register derived
from a subregister class of the original value.
https://invent.kde.org/qt/clang/llvm-project/-/commit/bb5f7a2a89e2662aaaef38717ab3a0352e77518a

Git commit 788b0d3f495ced424b64966dcebf982051ac1fc1 by Aaditya (on behalf of Christudasan Devadasan) on 31/07/2026 at 09:03..
[AMDGPU] Test precommit for subreg reload

This test currently fails due to insufficient
registers during allocation. Once the subreg
reload is implemented, it will begin to pass
as the partial reload help mitigate register
pressure.
https://invent.kde.org/qt/clang/llvm-project/-/commit/788b0d3f495ced424b64966dcebf982051ac1fc1

Git commit 841b269094897d289c9ef2e335c489489c969036 by Aaditya (on behalf of Christudasan Devadasan) on 31/07/2026 at 09:03..
compacted the virt-reg numbers
https://invent.kde.org/qt/clang/llvm-project/-/commit/841b269094897d289c9ef2e335c489489c969036

Git commit 1a848f4a7ae6bcb78e5320aebe55f879124b9e95 by Aaditya (on behalf of Christudasan Devadasan) on 31/07/2026 at 09:03..
[InlineSpiller][AMDGPU] Implement subreg reload during RA spill

Currently, when a virtual register is partially used, the
entire tuple is restored from the spilled location, even if
only a subset of its sub-registers is needed. This patch
introduces support for partial reloads by analyzing actual
register usage and restoring only the required sub-registers.
This improvement enhances register allocation efficiency,
particularly for cases involving tuple virtual registers.
For AMDGPU, this change brings considerable improvements
in workloads that involve matrix operations, large vectors,
and complex control flows.
https://invent.kde.org/qt/clang/llvm-project/-/commit/1a848f4a7ae6bcb78e5320aebe55f879124b9e95

Git commit 829c8e9360ece47f2ec76c67ee0f78306e2312cd by Aaditya on 31/07/2026 at 09:17..
Fix merge conflicts while rebasing on main.
https://invent.kde.org/qt/clang/llvm-project/-/commit/829c8e9360ece47f2ec76c67ee0f78306e2312cd