From: Jim Wilson <jim.wilson@linaro.org>
To: "gcc-patches@gcc.gnu.org" <gcc-patches@gcc.gnu.org>
Subject: [PATCH, AARCH64] make stdarg functions work with +nofp
Date: Sat, 23 May 2015 07:00:00 -0000 [thread overview]
Message-ID: <CABXYE2V+SwMbT5EsYM6hD0ENjwAhsZKAPTLvDuLxCNX9GAhLvg@mail.gmail.com> (raw)
[-- Attachment #1: Type: text/plain, Size: 851 bytes --]
The compiler currently ICEs when compiling a stdarg function with
+nofp, as reported in PR 66258.
The aarch64.md file disables FP instructions using TARGET_FLOAT, which
supports both -mgeneral-regs-only and +nofp. But there is code in
aarch64.c that checks TARGET_GENERAL_REGS_ONLY. This results in FP
instructions when using +nofp, The aarch64.c code needs to use
TARGET_FLOAT instead like the md file already does.
I can't meaningfully test this with a bootstrap, since the patch has
no effect unless I bootstrap with +nofp, and that will fail as gcc
contains floating point code.
The testsuite already has multiple stdarg tests, so there is no need
for another one.
I tested this by verifying I get the same results for some simple
testcasess with and without the patch, with and without using
-mgeneral-regs-only and -mcpu=cortex-a53+nofp.
[-- Attachment #2: a64-nofp-stdarg.patch --]
[-- Type: text/x-patch, Size: 2266 bytes --]
2015-05-22 Jim Wilson <jim.wilson@linaro.org>
PR target/66258
* config/aarch64/aarch64.c (aarch64_function_value_regno_p): Change
!TARGET_GENERAL_REGS_ONLY to TARGET_FLOAT.
(aarch64_secondary_reload): Likewise
(aarch64_expand_builtin_va_start): Change TARGET_GENERAL_REGS_ONLY
to !TARGET_FLOAT.
(aarch64_gimplify_va_arg_expr, aarch64_setup_incoming_varargs):
Likewise.
Index: config/aarch64/aarch64.c
===================================================================
--- config/aarch64/aarch64.c (revision 223590)
+++ config/aarch64/aarch64.c (working copy)
@@ -1666,7 +1666,7 @@ aarch64_function_value_regno_p (const un
/* Up to four fp/simd registers can return a function value, e.g. a
homogeneous floating-point aggregate having four members. */
if (regno >= V0_REGNUM && regno < V0_REGNUM + HA_MAX_NUM_FLDS)
- return !TARGET_GENERAL_REGS_ONLY;
+ return TARGET_FLOAT;
return false;
}
@@ -4783,7 +4783,7 @@ aarch64_secondary_reload (bool in_p ATTR
/* A TFmode or TImode memory access should be handled via an FP_REGS
because AArch64 has richer addressing modes for LDR/STR instructions
than LDP/STP instructions. */
- if (!TARGET_GENERAL_REGS_ONLY && rclass == GENERAL_REGS
+ if (TARGET_FLOAT && rclass == GENERAL_REGS
&& GET_MODE_SIZE (mode) == 16 && MEM_P (x))
return FP_REGS;
@@ -7571,7 +7571,7 @@ aarch64_expand_builtin_va_start (tree va
vr_save_area_size
= (NUM_FP_ARG_REGS - cum->aapcs_nvrn) * UNITS_PER_VREG;
- if (TARGET_GENERAL_REGS_ONLY)
+ if (!TARGET_FLOAT)
{
if (cum->aapcs_nvrn > 0)
sorry ("%qs and floating point or vector arguments",
@@ -7681,7 +7681,7 @@ aarch64_gimplify_va_arg_expr (tree valis
&is_ha))
{
/* TYPE passed in fp/simd registers. */
- if (TARGET_GENERAL_REGS_ONLY)
+ if (!TARGET_FLOAT)
sorry ("%qs and floating point or vector arguments",
"-mgeneral-regs-only");
@@ -7918,7 +7918,7 @@ aarch64_setup_incoming_varargs (cumulati
gr_saved = NUM_ARG_REGS - local_cum.aapcs_ncrn;
vr_saved = NUM_FP_ARG_REGS - local_cum.aapcs_nvrn;
- if (TARGET_GENERAL_REGS_ONLY)
+ if (!TARGET_FLOAT)
{
if (local_cum.aapcs_nvrn > 0)
sorry ("%qs and floating point or vector arguments",
next reply other threads:[~2015-05-22 23:24 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-05-23 7:00 Jim Wilson [this message]
2015-06-02 9:44 ` James Greenhalgh
2015-06-02 10:45 ` Kyrill Tkachov
2015-06-02 11:03 ` James Greenhalgh
2015-06-09 5:18 ` Jim Wilson
2015-06-16 8:48 ` James Greenhalgh
2015-06-16 21:05 ` Jim Wilson
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CABXYE2V+SwMbT5EsYM6hD0ENjwAhsZKAPTLvDuLxCNX9GAhLvg@mail.gmail.com \
--to=jim.wilson@linaro.org \
--cc=gcc-patches@gcc.gnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).