[llvm] [X86] LowerShift - if a vXi8 shift amount is small enough skip additional incremental shift stages (PR #174207)

via llvm-commits llvm-commits at lists.llvm.org
Fri Jan 2 05:10:05 PST 2026


github-actions[bot] wrote:

<!--LLVM CODE FORMAT COMMENT: {clang-format}-->


:warning: C/C++ code formatter, clang-format found issues in your code. :warning:

<details>
<summary>
You can test this locally with the following command:
</summary>

``````````bash
git-clang-format --diff origin/main HEAD --extensions cpp -- llvm/lib/Target/X86/X86ISelLowering.cpp --diff_from_common_commit
``````````

:warning:
The reproduction instructions above might return results for more than one PR
in a stack if you are using a stacked PR workflow. You can limit the results by
changing `origin/main` to the base branch/commit you want to compare against.
:warning:

</details>

<details>
<summary>
View the diff from clang-format here.
</summary>

``````````diff
diff --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp
index 25b557eeb..e49959444 100644
--- a/llvm/lib/Target/X86/X86ISelLowering.cpp
+++ b/llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -31250,10 +31250,11 @@ static SDValue LowerShift(SDValue Op, const X86Subtarget &Subtarget,
 
     // Turn 'a' into a mask suitable for VSELECT: a = a << 5;
     // We can safely do this using i16 shifts as we're only interested in
-    // the 3 lower bits of each byte. If the amount is even smaller, we can shift
-    // the mask further and skip additional incremental shifts.
+    // the 3 lower bits of each byte. If the amount is even smaller, we can
+    // shift the mask further and skip additional incremental shifts.
     Amt = DAG.getBitcast(ExtVT, Amt);
-    Amt = getTargetVShiftByConstNode(X86ISD::VSHLI, dl, ExtVT, Amt, 5 + MinLZ, DAG);
+    Amt = getTargetVShiftByConstNode(X86ISD::VSHLI, dl, ExtVT, Amt, 5 + MinLZ,
+                                     DAG);
     Amt = DAG.getBitcast(VT, Amt);
 
     if (Opc == ISD::SHL || Opc == ISD::SRL) {

``````````

</details>


https://github.com/llvm/llvm-project/pull/174207


More information about the llvm-commits mailing list