From: "Li, Pan2" <pan2.li@intel.com>
To: Kito Cheng <kito.cheng@gmail.com>
Cc: "juzhe.zhong@rivai.ai" <juzhe.zhong@rivai.ai>,
gcc-patches <gcc-patches@gcc.gnu.org>
Subject: RE: [PATCH v1] RISC-V: Bugfix ICE non-vector in TARGET_FUNCTION_VALUE_REGNO_P
Date: Sat, 13 Apr 2024 01:32:13 +0000 [thread overview]
Message-ID: <MW5PR11MB5908B18025FE7829BA2ACC25A90B2@MW5PR11MB5908.namprd11.prod.outlook.com> (raw)
In-Reply-To: <MW5PR11MB59084542647E6265F98E0A82A9042@MW5PR11MB5908.namprd11.prod.outlook.com>
Just completed the rv64imac test for fully regression test, there is NO increased failures.
For FP_RETURN, I think the ABI side somehow has some restrictions similar to
TARGET_HARD_FLOAT (of course as is). For example, the rv64imac cannot work
with lp64f/d, thus the FP_RETURN will be right REG because it is GP_RETURN if
the lp64 abi is given.
Unfortunately this is not working for v extension as we have no v in abi option.
How about we refine this part to TARGET_HARD_FLOAT after gcc-15
opens as the current implement looks like friable and implicit up to a point?
Pan
-----Original Message-----
From: Li, Pan2
Sent: Friday, April 12, 2024 6:58 PM
To: Kito Cheng <kito.cheng@gmail.com>
Cc: juzhe.zhong@rivai.ai; gcc-patches <gcc-patches@gcc.gnu.org>
Subject: RE: [PATCH v1] RISC-V: Bugfix ICE non-vector in TARGET_FUNCTION_VALUE_REGNO_P
Sure thing, the FP_RETURN only acts on ABI_xxx similar to below:
#define FP_RETURN (UNITS_PER_FP_ARG == 0 ? GP_RETURN : FP_ARG_FIRST)
I add some test for rv32/64imac option but don't cover all test cases without f/d extension, will have a try and keep you posted.
Pan
-----Original Message-----
From: Kito Cheng <kito.cheng@gmail.com>
Sent: Friday, April 12, 2024 4:56 PM
To: Li, Pan2 <pan2.li@intel.com>
Cc: juzhe.zhong@rivai.ai; gcc-patches <gcc-patches@gcc.gnu.org>
Subject: Re: [PATCH v1] RISC-V: Bugfix ICE non-vector in TARGET_FUNCTION_VALUE_REGNO_P
Does FP reg also need gurared with TARGET_HARD_FLOAT? could you try to
compile that case without F?
On Fri, Apr 12, 2024 at 2:19 PM Li, Pan2 <pan2.li@intel.com> wrote:
>
> Committed, thanks Juzhe.
>
>
>
> Pan
>
>
>
> From: juzhe.zhong@rivai.ai <juzhe.zhong@rivai.ai>
> Sent: Friday, April 12, 2024 2:11 PM
> To: Li, Pan2 <pan2.li@intel.com>; gcc-patches <gcc-patches@gcc.gnu.org>
> Cc: kito.cheng <kito.cheng@gmail.com>; Li, Pan2 <pan2.li@intel.com>
> Subject: Re: [PATCH v1] RISC-V: Bugfix ICE non-vector in TARGET_FUNCTION_VALUE_REGNO_P
>
>
>
> LGTM。
>
>
>
> ________________________________
>
> juzhe.zhong@rivai.ai
>
>
>
> From: pan2.li
>
> Date: 2024-04-12 14:08
>
> To: gcc-patches
>
> CC: juzhe.zhong; kito.cheng; Pan Li
>
> Subject: [PATCH v1] RISC-V: Bugfix ICE non-vector in TARGET_FUNCTION_VALUE_REGNO_P
>
> From: Pan Li <pan2.li@intel.com>
>
>
>
> This patch would like to fix one ICE when vector is not enabled
>
> in hook TARGET_FUNCTION_VALUE_REGNO_P implementation. The vector
>
> regno is available if and only if the TARGET_VECTOR is true. The
>
> previous implement missed this condition and then result in ICE
>
> when rv64gc build option without vector.
>
>
>
> PR target/114639
>
>
>
> The below test suite is passed for this patch.
>
>
>
> * The rv64gcv fully regression tests.
>
> * The rv64gc fully regression tests.
>
>
>
> gcc/ChangeLog:
>
>
>
> * config/riscv/riscv.cc (riscv_function_value_regno_p): Add
>
> TARGET_VECTOR predicate for V_RETURN regno.
>
>
>
> gcc/testsuite/ChangeLog:
>
>
>
> * gcc.target/riscv/pr114639-1.c: New test.
>
> * gcc.target/riscv/pr114639-2.c: New test.
>
> * gcc.target/riscv/pr114639-3.c: New test.
>
> * gcc.target/riscv/pr114639-4.c: New test.
>
>
>
> Signed-off-by: Pan Li <pan2.li@intel.com>
>
> ---
>
> gcc/config/riscv/riscv.cc | 2 +-
>
> gcc/testsuite/gcc.target/riscv/pr114639-1.c | 11 +++++++++++
>
> gcc/testsuite/gcc.target/riscv/pr114639-2.c | 11 +++++++++++
>
> gcc/testsuite/gcc.target/riscv/pr114639-3.c | 11 +++++++++++
>
> gcc/testsuite/gcc.target/riscv/pr114639-4.c | 11 +++++++++++
>
> 5 files changed, 45 insertions(+), 1 deletion(-)
>
> create mode 100644 gcc/testsuite/gcc.target/riscv/pr114639-1.c
>
> create mode 100644 gcc/testsuite/gcc.target/riscv/pr114639-2.c
>
> create mode 100644 gcc/testsuite/gcc.target/riscv/pr114639-3.c
>
> create mode 100644 gcc/testsuite/gcc.target/riscv/pr114639-4.c
>
>
>
> diff --git a/gcc/config/riscv/riscv.cc b/gcc/config/riscv/riscv.cc
>
> index 91f017dd52a..e5f00806bb9 100644
>
> --- a/gcc/config/riscv/riscv.cc
>
> +++ b/gcc/config/riscv/riscv.cc
>
> @@ -11008,7 +11008,7 @@ riscv_function_value_regno_p (const unsigned regno)
>
> if (FP_RETURN_FIRST <= regno && regno <= FP_RETURN_LAST)
>
> return true;
>
> - if (regno == V_RETURN)
>
> + if (TARGET_VECTOR && regno == V_RETURN)
>
> return true;
>
> return false;
>
> diff --git a/gcc/testsuite/gcc.target/riscv/pr114639-1.c b/gcc/testsuite/gcc.target/riscv/pr114639-1.c
>
> new file mode 100644
>
> index 00000000000..f41723193a4
>
> --- /dev/null
>
> +++ b/gcc/testsuite/gcc.target/riscv/pr114639-1.c
>
> @@ -0,0 +1,11 @@
>
> +/* Test that we do not have ice when compile */
>
> +/* { dg-do compile } */
>
> +/* { dg-options "-march=rv64gc -mabi=lp64d -std=gnu89 -O3" } */
>
> +
>
> +g (a, b) {}
>
> +
>
> +f (xx)
>
> + void* xx;
>
> +{
>
> + __builtin_apply ((void*)g, xx, 200);
>
> +}
>
> diff --git a/gcc/testsuite/gcc.target/riscv/pr114639-2.c b/gcc/testsuite/gcc.target/riscv/pr114639-2.c
>
> new file mode 100644
>
> index 00000000000..0c402c4b254
>
> --- /dev/null
>
> +++ b/gcc/testsuite/gcc.target/riscv/pr114639-2.c
>
> @@ -0,0 +1,11 @@
>
> +/* Test that we do not have ice when compile */
>
> +/* { dg-do compile } */
>
> +/* { dg-options "-march=rv64imac -mabi=lp64 -std=gnu89 -O3" } */
>
> +
>
> +g (a, b) {}
>
> +
>
> +f (xx)
>
> + void* xx;
>
> +{
>
> + __builtin_apply ((void*)g, xx, 200);
>
> +}
>
> diff --git a/gcc/testsuite/gcc.target/riscv/pr114639-3.c b/gcc/testsuite/gcc.target/riscv/pr114639-3.c
>
> new file mode 100644
>
> index 00000000000..ffb0d6d162d
>
> --- /dev/null
>
> +++ b/gcc/testsuite/gcc.target/riscv/pr114639-3.c
>
> @@ -0,0 +1,11 @@
>
> +/* Test that we do not have ice when compile */
>
> +/* { dg-do compile } */
>
> +/* { dg-options "-march=rv32gc -mabi=ilp32d -std=gnu89 -O3" } */
>
> +
>
> +g (a, b) {}
>
> +
>
> +f (xx)
>
> + void* xx;
>
> +{
>
> + __builtin_apply ((void*)g, xx, 200);
>
> +}
>
> diff --git a/gcc/testsuite/gcc.target/riscv/pr114639-4.c b/gcc/testsuite/gcc.target/riscv/pr114639-4.c
>
> new file mode 100644
>
> index 00000000000..a6e229101ef
>
> --- /dev/null
>
> +++ b/gcc/testsuite/gcc.target/riscv/pr114639-4.c
>
> @@ -0,0 +1,11 @@
>
> +/* Test that we do not have ice when compile */
>
> +/* { dg-do compile } */
>
> +/* { dg-options "-march=rv32imac -mabi=ilp32 -std=gnu89 -O3" } */
>
> +
>
> +g (a, b) {}
>
> +
>
> +f (xx)
>
> + void* xx;
>
> +{
>
> + __builtin_apply ((void*)g, xx, 200);
>
> +}
>
> --
>
> 2.34.1
>
>
>
>
prev parent reply other threads:[~2024-04-13 1:32 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-04-12 6:08 pan2.li
2024-04-12 6:11 ` juzhe.zhong
2024-04-12 6:18 ` Li, Pan2
2024-04-12 8:55 ` Kito Cheng
2024-04-12 10:57 ` Li, Pan2
2024-04-13 1:32 ` Li, Pan2 [this message]
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=MW5PR11MB5908B18025FE7829BA2ACC25A90B2@MW5PR11MB5908.namprd11.prod.outlook.com \
--to=pan2.li@intel.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=juzhe.zhong@rivai.ai \
--cc=kito.cheng@gmail.com \
/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).