File aarch64-sve-powf.patch of Package glibc

From e34453cd6a8c592c325756ff3c7ac0afd3975cb4 Mon Sep 17 00:00:00 2001
From: Pierre Blanchard <pierre.blanchard@arm.com>
Date: Wed, 20 Aug 2025 17:41:50 +0000
Subject: [PATCH] AArch64: Fix SVE powf routine [BZ #33299]

Fix a bug in predicate logic introduced in last change.
A slight performance improvement from relying on all true
predicates during conversion from single to double.
This fixes BZ #33299.

Reviewed-by: Wilco Dijkstra  <Wilco.Dijkstra@arm.com>
(cherry picked from commit aac077645a645bba0d67f3250e82017c539d0f4b)
---
 sysdeps/aarch64/fpu/powf_sve.c | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

diff --git a/sysdeps/aarch64/fpu/powf_sve.c b/sysdeps/aarch64/fpu/powf_sve.c
index 7046990aa1..65e9bd29d9 100644
--- a/sysdeps/aarch64/fpu/powf_sve.c
+++ b/sysdeps/aarch64/fpu/powf_sve.c
@@ -223,15 +223,15 @@ sv_powf_core (const svbool_t pg, svuint32_t i, svuint32_t iz, svint32_t k,
   const svbool_t ptrue = svptrue_b64 ();
 
   /* Unpack and promote input vectors (pg, y, z, i, k and sign_bias) into two
-   * in order to perform core computation in double precision.  */
+     in order to perform core computation in double precision.  */
   const svbool_t pg_lo = svunpklo (pg);
   const svbool_t pg_hi = svunpkhi (pg);
-  svfloat64_t y_lo
-      = svcvt_f64_x (pg, svreinterpret_f32 (svunpklo (svreinterpret_u32 (y))));
-  svfloat64_t y_hi
-      = svcvt_f64_x (pg, svreinterpret_f32 (svunpkhi (svreinterpret_u32 (y))));
-  svfloat64_t z_lo = svcvt_f64_x (pg, svreinterpret_f32 (svunpklo (iz)));
-  svfloat64_t z_hi = svcvt_f64_x (pg, svreinterpret_f32 (svunpkhi (iz)));
+  svfloat64_t y_lo = svcvt_f64_x (
+      ptrue, svreinterpret_f32 (svunpklo (svreinterpret_u32 (y))));
+  svfloat64_t y_hi = svcvt_f64_x (
+      ptrue, svreinterpret_f32 (svunpkhi (svreinterpret_u32 (y))));
+  svfloat64_t z_lo = svcvt_f64_x (ptrue, svreinterpret_f32 (svunpklo (iz)));
+  svfloat64_t z_hi = svcvt_f64_x (ptrue, svreinterpret_f32 (svunpkhi (iz)));
   svuint64_t i_lo = svunpklo (i);
   svuint64_t i_hi = svunpkhi (i);
   svint64_t k_lo = svunpklo (k);
@@ -312,7 +312,7 @@ svfloat32_t SV_NAME_F2 (pow) (svfloat32_t x, svfloat32_t y, const svbool_t pg)
 			 (23 - V_POWF_EXP2_TABLE_BITS));
 
   /* Compute core in extended precision and return intermediate ylogx results
-   * to handle cases of underflow and underflow in exp.  */
+     to handle cases of underflow and overflow in exp.  */
   svfloat32_t ylogx;
   svfloat32_t ret
       = sv_powf_core (yint_or_xpos, i, iz, k, y, sign_bias, &ylogx, d);
-- 
2.51.0