[all-commits] [llvm/llvm-project] 5d31c0: [lldb] Fix switch case for i386 in COFF GetModuleS...

Fri Sep 25 01:24:05 PDT 2026

  Branch: refs/heads/release/23.x
  Home:   https://github.com/llvm/llvm-project
  Commit: 5d31c0334b0378d560f6a15705ec65ff8772e5b4
      https://github.com/llvm/llvm-project/commit/5d31c0334b0378d560f6a15705ec65ff8772e5b4
  Author: David Spickett <david.spickett at arm.com>
  Date:   2026-09-25 (Fri, 25 Sep 2026)

  Changed paths:
    M lldb/source/Plugins/ObjectFile/COFF/ObjectFileCOFF.cpp

  Log Message:
  -----------
  [lldb] Fix switch case for i386 in COFF GetModuleSpecifications (#222030)

08c94c0ac3b503dd4971da8935001aba01918688 / #188276 accidentally removed
the "case:" for i386.

(cherry picked from commit 443d419e798b37a3733b12db7859c5155407c6d8)

  Commit: ff3d38593e6de5e0fd909bdcc52e9a5584f49215
      https://github.com/llvm/llvm-project/commit/ff3d38593e6de5e0fd909bdcc52e9a5584f49215
  Author: yingopq <115543042+yingopq at users.noreply.github.com>
  Date:   2026-09-25 (Fri, 25 Sep 2026)

  Changed paths:
    M llvm/lib/Target/Mips/MipsAsmPrinter.cpp
    M llvm/lib/Target/Mips/MipsAsmPrinter.h
    A llvm/test/CodeGen/Mips/jalr-no-mangle.ll

  Log Message:
  -----------
  [Mips] Strip \x01 no-mangle prefix from R_MIPS_JALR symbol name (#219427)

R_MIPS_JALR relocation did not strip the \x01 no-mangle prefix, causing
linker to see two different symbols: my_target_sym and
\x01my_target_sym.

Use Mangler::getNameWithPrefix() to process the symbol name.

Fix #207470.

(cherry picked from commit f796b25647e378a02482ef383f2db45804d8fb72)

  Commit: e24fcb6944108ec6013eaf466101b1bcb45304d5
      https://github.com/llvm/llvm-project/commit/e24fcb6944108ec6013eaf466101b1bcb45304d5
  Author: David Green <david.green at arm.com>
  Date:   2026-09-25 (Fri, 25 Sep 2026)

  Changed paths:
    M llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
    M llvm/test/CodeGen/AArch64/bcmp.ll
    M llvm/test/CodeGen/AArch64/i128-imm-compare-ccmp.ll

  Log Message:
  -----------
  [AArch64] Guard against large types in performOrXorChainCombine (#222146)

Larger type sizes will be legalized to i64, and will hit a crash in
checking constants.

Fixes #222089

(cherry picked from commit 58c46bae21181fb4240cf964461bea738d96d50e)

  Commit: 28a01b03e11ffeb3012677f12a8e18d55ba1847d
      https://github.com/llvm/llvm-project/commit/28a01b03e11ffeb3012677f12a8e18d55ba1847d
  Author: Nerixyz <nerixdev at outlook.de>
  Date:   2026-09-25 (Fri, 25 Sep 2026)

  Changed paths:
    M lldb/include/lldb/API/SBFile.h
    M lldb/source/API/SBFile.cpp
    M lldb/source/Plugins/ScriptInterpreter/Python/PythonDataObjects.cpp
    M lldb/source/Plugins/ScriptInterpreter/Python/PythonDataObjects.h
    M lldb/source/Plugins/ScriptInterpreter/Python/ScriptInterpreterPython.cpp
    M lldb/tools/lldb-dap/DAP.cpp

  Log Message:
  -----------
  [lldb][Windows] Translate file descriptors between CRT instances (#217126)

The official LLVM and lldb builds use `LLVM_ENABLE_RPMALLOC`. This
changes the default CRT allocator to rpmalloc for performance reasons.
In doing so, it also changes the CRT linkage from dynamic (`/MD`) to
static (`/MT`).
Since the CRT is the one managing file descriptors on Windows, its
universe is now scoped to a module. For lldb, this means we have three
fd-spaces: the driver (lldb/lldb-dap), liblldb, and Python.
We usually pass files around as file descriptors. This breaks when we do
it between modules.

Fortunately, the C runtime provides us with helpers to translate between
the process scoped HANDLEs and file descriptors. This way, we can
allocate an fd for a handle passed across modules. For Python, there's
the [`msvcrt`](https://docs.python.org/3/library/msvcrt.html) library we
can use to translate.

Fixes #216679.

(cherry picked from commit 1ca6698be7708121f987d0acf896c7dc4cf63a51)

  Commit: 5f2881edaf60c7f793ea87d5ad0305791af50831
      https://github.com/llvm/llvm-project/commit/5f2881edaf60c7f793ea87d5ad0305791af50831
  Author: Nerixyz <nerixdev at outlook.de>
  Date:   2026-09-25 (Fri, 25 Sep 2026)

  Changed paths:
    M lldb/include/lldb/API/SBFile.h
    M lldb/source/API/SBFile.cpp
    M lldb/source/Plugins/ScriptInterpreter/Python/PythonDataObjects.cpp

  Log Message:
  -----------
  [lldb][Windows] Remove use of unstable ABI and fix SB API visibility (#222665)

Addresses the post-commit review comments from the original PR
(https://github.com/llvm/llvm-project/pull/217126#pullrequestreview-5162589053).

- `SBFile::OpenFdFromHandle` is now visible on all platforms (just
`#ifndef SWIG`) and returns -1 by default.
- Removed non-stable `PyLong_Check` in favor of `PyErr_Occurred` and
`PyErr_Clear`.

(cherry picked from commit ba1687b72476eb368641b9a9f565a38cf41b7393)

  Commit: 3273cdaa5ea1188a6dc630039f0be90d8d8084b6
      https://github.com/llvm/llvm-project/commit/3273cdaa5ea1188a6dc630039f0be90d8d8084b6
  Author: David Green <david.green at arm.com>
  Date:   2026-09-25 (Fri, 25 Sep 2026)

  Changed paths:
    M llvm/test/CodeGen/AArch64/fptrunc.ll
    A llvm/test/TableGen/GlobalISelEmitter/MatchTableOptimizerTypes.td
    M llvm/utils/TableGen/Common/GlobalISel/MatchTable/Matchers.cpp

  Log Message:
  -----------
  [GlobalISel][AArch64] Do not remove dst type predicates from pattern children nodes. (#221576)

Given an instruction with a non-concrete type (G_CONCAT_VECTOR), with a
child pattern instruction with a concrete type (G_FPTRUNC), the concrete
type check for the child would incorrectly be removed. This patch makes
sure they remain. This makes the match table a little larger, but many
should be able to be optimized away (providing that the two operands
with the same type can be matched up).

Fixes #221458

(cherry picked from commit 09be00c800cf9ae03a2c8fbdc215c259ac39f30f)

  Commit: 6b34061eb9d4c19441a1444ba5c7207d7703ad1d
      https://github.com/llvm/llvm-project/commit/6b34061eb9d4c19441a1444ba5c7207d7703ad1d
  Author: David Tellenbach <dtellenbach at apple.com>
  Date:   2026-09-25 (Fri, 25 Sep 2026)

  Changed paths:
    M llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
    A llvm/test/CodeGen/AArch64/sme-za-non-sibling-tailcall.ll
    M llvm/test/CodeGen/AArch64/sme-za-tailcall-fpdiff-align.ll

  Log Message:
  -----------
  [AArch64] Don't emit stack restore for SME ZA non-sibling tail calls (#224721)

A tail call from a function with live ZA state was previously prevented
from being lowered as a sibling call to have a CALLSEQ_START to glue
INOUT_ZA_USE to. This led to bogus stack restores.

We now drop the INOUT_ZA_USE marker on tail calls and can thus lower as
sibling calls if needed to take advantage of the existing correct stack
restore behavior.

This is safe to do since tail calls use TCRETURN and the MachineSMEABI
pass requires live ZA state to stay live across returns anyway.

(cherry picked from commit ad695d46e4535f953558705541b71c2d938fb72b)

  Commit: 4db63eeac73fe88f2f6f0deba04b1a84d5e9f63d
      https://github.com/llvm/llvm-project/commit/4db63eeac73fe88f2f6f0deba04b1a84d5e9f63d
  Author: Nemanja Ivanovic <nemanja.i.llvm at gmail.com>
  Date:   2026-09-25 (Fri, 25 Sep 2026)

  Changed paths:
    M llvm/lib/Target/RISCV/RISCVFrameLowering.cpp
    M llvm/lib/Target/RISCV/RISCVPushPopOptimizer.cpp
    A llvm/test/CodeGen/RISCV/shadow-stack-zcmp.ll
    M llvm/test/CodeGen/RISCV/shadowcallstack.ll

  Log Message:
  -----------
  [RISC-V] Do not emit cm.popret[z] with zicfiss (#196267)

When emitting shadow call stack protection instructions, the push/pop
optimization needs to be turned off because an sspopchk before a
cm.popret[z] is guaranteed to fail in a non-leaf function. In addition,
the sspopchk must be emitted after a cm.pop so that the ra has the
correct value when the check is performed.

Fixes: https://github.com/llvm/llvm-project/issues/196261

Co-authored-by: Nemanja Ivanovic <nemanja at synopsys.com>
(cherry picked from commit 255162ae0ebe805df151c5c4b48e1a47a5dd74f0)

  Commit: c9376624635f07153a643be1178a66d6c25ca8e7
      https://github.com/llvm/llvm-project/commit/c9376624635f07153a643be1178a66d6c25ca8e7
  Author: Nemanja Ivanovic <nemanja.i.llvm at gmail.com>
  Date:   2026-09-25 (Fri, 25 Sep 2026)

  Changed paths:
    M llvm/lib/Target/RISCV/RISCVFrameLowering.cpp
    A llvm/test/CodeGen/RISCV/epilog-crash.ll

  Log Message:
  -----------
  [RISC-V] Fix assert after 255162a (#212791)

The iterator passed-in may point to the end of the block which causes an
assertion failure when attempting to inspect the MI it points to. Guard
against this.

(cherry picked from commit f24da9d03ba5d25c523e050da2ca3c4060ca23d5)

Compare: https://github.com/llvm/llvm-project/compare/898bafa6c855...c9376624635f

To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications