public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: Lulu Cheng <chenglulu@loongson.cn>
To: WANG Xuerui <i@xen0n.name>, Xi Ruoyao <xry111@xry111.site>,
	gcc-patches@gcc.gnu.org,
	"zhangdandan@loongson.cn" <zhangdandan@loongson.cn>
Cc: Chenghua Xu <xuchenghua@loongson.cn>, Revy <rabenda.cn@gmail.com>
Subject: Re: [PATCH] LoongArch: Fix multiarch tuple canonization
Date: Fri, 17 Feb 2023 09:41:19 +0800	[thread overview]
Message-ID: <5b2f1ce1-6acc-0388-6ca3-30b7227103d1@loongson.cn> (raw)
In-Reply-To: <a39aa0fc-234d-fa40-53ec-0a06f0a31c94@xen0n.name>

Hi,

在 2023/2/15 下午6:42, WANG Xuerui 写道:
> Hi,
>
> On 2023/2/13 18:38, Xi Ruoyao wrote:
>> Multiarch tuple will be coded in file or directory names in
>> multiarch-aware distros, so one ABI should have only one multiarch
>> tuple.  For example, "--target=loongarch64-linux-gnu --with-abi=lp64s"
>> and "--target=loongarch64-linux-gnusf" should both set multiarch tuple
>> to "loongarch64-linux-gnusf".  Before this commit,
>> "--target=loongarch64-linux-gnu --with-abi=lp64s --disable-multilib"
>> will produce wrong result (loongarch64-linux-gnu).
>>
>> A recent LoongArch psABI revision mandates "loongarch64-linux-gnu" to be
>> used for -mabi=lp64d (instead of "loongarch64-linux-gnuf64") for some
>> non-technical reason [1].  Note that we cannot make
>> "loongarch64-linux-gnuf64" an alias for "loongarch64-linux-gnu" because
>> to implement such an alias, we must create thousands of symlinks in the
>> distro and doing so would be completely unpractical.  This commit also
>> aligns GCC with the revision.
>>
>> Tested by building cross compilers with --enable-multiarch and multiple
>> combinations of --target=loongarch64-linux-gnu*, --with-abi=lp64{s,f,d},
>> and --{enable,disable}-multilib; and run "xgcc --print-multiarch" then
>> manually verify the result with eyesight.
>>
>> Ok for trunk and backport to releases/gcc-12?
>>
>> [1]: https://github.com/loongson/LoongArch-Documentation/pull/80
>>
>> gcc/ChangeLog:
>>
>>     * config.gcc (triplet_abi): Set its value based on $with_abi,
>>     instead of $target.
>>     (la_canonical_triplet): Set it after $triplet_abi is set
>>     correctly.
>>     * config/loongarch/t-linux (MULTILIB_OSDIRNAMES): Make the
>>     multiarch tuple for lp64d "loongarch64-linux-gnu" (without
>>     "f64" suffix).
>> ---
>>   gcc/config.gcc               | 14 +++++++-------
>>   gcc/config/loongarch/t-linux |  2 +-
>>   2 files changed, 8 insertions(+), 8 deletions(-)
>>
>> diff --git a/gcc/config.gcc b/gcc/config.gcc
>> index 067720ac795..c070e6ecd2e 100644
>> --- a/gcc/config.gcc
>> +++ b/gcc/config.gcc
>> @@ -4889,20 +4889,16 @@ case "${target}" in
>>           case ${target} in
>>           loongarch64-*-*-*f64)
>>               abi_pattern="lp64d"
>> -            triplet_abi="f64"
>>               ;;
>>           loongarch64-*-*-*f32)
>>               abi_pattern="lp64f"
>> -            triplet_abi="f32"
>>               ;;
>>           loongarch64-*-*-*sf)
>>               abi_pattern="lp64s"
>> -            triplet_abi="sf"
>>               ;;
>>           loongarch64-*-*-*)
>>               abi_pattern="lp64[dfs]"
>>               abi_default="lp64d"
>> -            triplet_abi=""
>>               ;;
>>           *)
>>               echo "Unsupported target ${target}." 1>&2
>> @@ -4923,9 +4919,6 @@ case "${target}" in
>>                 ;;
>>           esac
>>   - la_canonical_triplet="loongarch64-${triplet_os}${triplet_abi}"
>> -
>> -
>>           # Perform initial sanity checks on --with-* options.
>>           case ${with_arch} in
>>           "" | loongarch64 | la464) ;; # OK, append here.
>> @@ -4996,6 +4989,13 @@ case "${target}" in
>>               ;;
>>           esac
>>   +        case ${with_abi} in
>> +          "lp64d") triplet_abi="";;
>> +          "lp64f") triplet_abi="f32";;
>> +          "lp64s") triplet_abi="sf";;
>> +        esac
>> + la_canonical_triplet="loongarch64-${triplet_os}${triplet_abi}"
>> +
>>           # Set default value for with_abiext (internal)
>>           case ${with_abiext} in
>>           "")
>> diff --git a/gcc/config/loongarch/t-linux b/gcc/config/loongarch/t-linux
>> index 131c45fdced..e40da179203 100644
>> --- a/gcc/config/loongarch/t-linux
>> +++ b/gcc/config/loongarch/t-linux
>> @@ -40,7 +40,7 @@ ifeq ($(filter LA_DISABLE_MULTILIB,$(tm_defines)),)
>>         MULTILIB_OSDIRNAMES = \
>>         mabi.lp64d=../lib64$\
>> -      $(call if_multiarch,:loongarch64-linux-gnuf64)
>> +      $(call if_multiarch,:loongarch64-linux-gnu)
>>         MULTILIB_OSDIRNAMES += \
>>         mabi.lp64f=../lib64/f32$\
>
> Thanks for the quick patch; however Revy told me offline yesterday 
> that this might conflict with things Debian side once this gets 
> merged. He may have more details to share.
>
> Adding him to CC -- you could keep him CC-ed on future changes that 
> may impact distro packaging.

Thank you for your feedback.

This modification plan is determined by the operating system group. If 
there is any problem, you can describe it clearly.

If there is no problem, we will combine this patch.

Thanks!


  reply	other threads:[~2023-02-17  1:41 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-02-13 10:38 Xi Ruoyao
2023-02-14  3:32 ` Lulu Cheng
2023-02-15  9:32   ` Yujie Yang
2023-02-15 10:42 ` WANG Xuerui
2023-02-17  1:41   ` Lulu Cheng [this message]
2023-02-18  8:59     ` [pushed][PATCH] " Lulu Cheng

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=5b2f1ce1-6acc-0388-6ca3-30b7227103d1@loongson.cn \
    --to=chenglulu@loongson.cn \
    --cc=gcc-patches@gcc.gnu.org \
    --cc=i@xen0n.name \
    --cc=rabenda.cn@gmail.com \
    --cc=xry111@xry111.site \
    --cc=xuchenghua@loongson.cn \
    --cc=zhangdandan@loongson.cn \
    /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).