diff options
| author | BtbN <btbn@btbn.de> | 2024-08-10 16:48:55 +0200 |
|---|---|---|
| committer | Chris Packham <judge.packham@gmail.com> | 2024-08-22 13:02:32 +1200 |
| commit | 5595edc370d8146ca3bbb3052dde48aceaff4970 (patch) | |
| tree | a40d018ea025224935e82b67e9ee90b10bcf4313 /packages/gcc/14.2.0/0016-Revert-one-of-the-force_subreg-changes.patch | |
| parent | ed12fa68402f58e171a6f79500f73f4781fdc9e5 (diff) | |
| download | crosstool-ng-5595edc370d8146ca3bbb3052dde48aceaff4970.tar.gz crosstool-ng-5595edc370d8146ca3bbb3052dde48aceaff4970.tar.bz2 crosstool-ng-5595edc370d8146ca3bbb3052dde48aceaff4970.zip | |
gcc-14.2: Fix ICE on aarch64
See https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115464 for issue
description and list of patches to backport.
Signed-off-by: BtbN <btbn@btbn.de>
Diffstat (limited to 'packages/gcc/14.2.0/0016-Revert-one-of-the-force_subreg-changes.patch')
| -rw-r--r-- | packages/gcc/14.2.0/0016-Revert-one-of-the-force_subreg-changes.patch | 46 |
1 files changed, 46 insertions, 0 deletions
diff --git a/packages/gcc/14.2.0/0016-Revert-one-of-the-force_subreg-changes.patch b/packages/gcc/14.2.0/0016-Revert-one-of-the-force_subreg-changes.patch new file mode 100644 index 00000000..629cfbda --- /dev/null +++ b/packages/gcc/14.2.0/0016-Revert-one-of-the-force_subreg-changes.patch @@ -0,0 +1,46 @@ +From 5468439a1f987b7d801c6c76d6c989e57af8916a Mon Sep 17 00:00:00 2001 +From: Richard Sandiford <richard.sandiford@arm.com> +Date: Tue, 25 Jun 2024 09:41:21 +0100 +Subject: [PATCH 16/16] Revert one of the force_subreg changes + +One of the changes in g:d4047da6a070175aae7121c739d1cad6b08ff4b2 +caused a regression in ft32-elf; see: + + https://gcc.gnu.org/pipermail/gcc-patches/2024-June/655418.html + +for details. This change was different from the others in that the +original call was to simplify_subreg rather than simplify_lowpart_subreg. +The old code would therefore go on to do the force_reg for more cases +than the new code would. + +gcc/ + * expmed.cc (store_bit_field_using_insv): Revert earlier change + to use force_subreg instead of simplify_gen_subreg. + +(cherry picked from commit b694bf417cdd7d0a4d78e9927bab6bc202b7df6c) +--- + gcc/expmed.cc | 8 +++++++- + 1 file changed, 7 insertions(+), 1 deletion(-) + +diff --git a/gcc/expmed.cc b/gcc/expmed.cc +index bd190722de6..85ec2614a3f 100644 +--- a/gcc/expmed.cc ++++ b/gcc/expmed.cc +@@ -695,7 +695,13 @@ store_bit_field_using_insv (const extraction_insn *insv, rtx op0, + if we must narrow it, be sure we do it correctly. */ + + if (GET_MODE_SIZE (value_mode) < GET_MODE_SIZE (op_mode)) +- tmp = force_subreg (op_mode, value1, value_mode, 0); ++ { ++ tmp = simplify_subreg (op_mode, value1, value_mode, 0); ++ if (! tmp) ++ tmp = simplify_gen_subreg (op_mode, ++ force_reg (value_mode, value1), ++ value_mode, 0); ++ } + else + { + tmp = gen_lowpart_if_possible (op_mode, value1); +-- +2.44.2 + |