Re: [PATCH 1/1] RISC-V: Fix canonical extension order (K and J)

Re: [PATCH 1/1] RISC-V: Fix canonical extension order (K and J)

[Tsukasa OI]

On 2022/05/19 12:40, Nelson Chu wrote:
> Seems like gcc and llvm have already committed this patch, so LGTM, committed.

Sorry, the same change is applied to LLVM but not yet on GCC (because I
forgot to add "Signed-off-by" line).  I sent PATCH v2 to gcc-patches
today so that would be okay.  On PATCH v2, I made the same change to
gcc/config/riscv/arch-canonicalize script (not just
gcc/common/config/riscv/riscv-common.cc).

<https://gcc.gnu.org/pipermail/gcc-patches/2022-May/595373.html>

Thanks,
Tsukasa

> 
> Thanks
> Nelson
> 
> On Mon, Apr 25, 2022 at 11:53 AM Palmer Dabbelt <palmer@dabbelt.com> wrote:
>>
>> On Mon, 28 Mar 2022 17:12:55 PDT (-0700), Palmer Dabbelt wrote:
>>> On Mon, 28 Mar 2022 06:12:01 PDT (-0700), binutils@sourceware.org wrote:
>>>> This commit fixes canonical extension order to follow the RISC-V ISA
>>>> Manual draft-20210402-1271737 or later.
>>>>
>>>> bfd/ChangeLog:
>>>>
>>>>      * elfxx-riscv.c (riscv_recognized_prefixed_ext): Fix "K" extension
>>>>      prefix to be placed before "J".
>>>> ---
>>>>  bfd/elfxx-riscv.c | 2 +-
>>>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>>>
>>>> diff --git a/bfd/elfxx-riscv.c b/bfd/elfxx-riscv.c
>>>> index cb2cc146c04..1219a7b44d4 100644
>>>> --- a/bfd/elfxx-riscv.c
>>>> +++ b/bfd/elfxx-riscv.c
>>>> @@ -1338,7 +1338,7 @@ riscv_recognized_prefixed_ext (const char *ext)
>>>>  }
>>>>
>>>>  /* Canonical order for single letter extensions.  */
>>>> -static const char riscv_ext_canonical_order[] = "eigmafdqlcbjktpvn";
>>>> +static const char riscv_ext_canonical_order[] = "eigmafdqlcbkjtpvn";
>>>>
>>>>  /* Array is used to compare the orders of standard extensions quickly.  */
>>>>  static int riscv_ext_order[26] = {0};
>>>
>>> Looks like this was just a bug in binutils: K went from being
>>> unspecified to specified in 271737 ("Define canonical location of K
>>> extension in ISA string"), thus it was never allowed at that other bit
>>> position.
>>>
>>> It looks like GCC also has this wrong, which sort of doubles the
>>> headache: now we've got this odd coupling between the GCC version and
>>> binutils version.  I'm not sure what the right thing is to do here:
>>> certainly rejecting the valid ISA string should be fixed, but I think we
>>> might need to accept the invalid one for compatibility reasons.  That'll
>>> be a headache to implement, though, so I'm not sure it's worth it.
>>>
>>> Maybe someone has a clever solution to this one?
>>
>> After seeing the GCC patch go by, I think the clever solution here is to
>> just say that we never accepted any J stuff in the first place so it's
>> not a compatibility break.
> 

Re: [PATCH 1/1] RISC-V: Fix canonical extension order (K and J)

[Kito Cheng]

> > and so it doesn't make
> > sense to mandate any particular ordering.
>
> No. It affects Z* extension ordering...

+1, we really need the order in ISA spec so that we could know the
canonical order for z* exts.

Re: [PATCH 1/1] RISC-V: Fix canonical extension order (K and J)

[Tsukasa OI]

Hello,

> Neither K nor J is an extension that exists,

That is correct.

> and so it doesn't make
> sense to mandate any particular ordering.

No. It affects Z* extension ordering...


On 2022/04/24 14:36, Andrew Waterman wrote:
> Neither K nor J is an extension that exists, and so it doesn't make
> sense to mandate any particular ordering.  The better change would be
> to delete the letters `k' and `j' from that string, so that we aren't
> enforcing constraints that don't serve a useful purpose.
> 
> cf. https://github.com/riscv/riscv-isa-manual/commit/f5f9c27010b69a015958ffebe1ac5a34f8776dff

Wait... so, you make constraints for existing single-letters (Zi -> Zv)
but not for non-existing single-letters? (Zk -> Zj, Zj -> Zk) anymore?

That's completely unexpected move but also makes sense.

Let me check your intentions and details: do we need to place Z[CH]*
extensions without single-letter extension [CH] after all existing ones
(like Zv*)? Or, Z[CH]* extensions without single-letter extension [CH]
have no constraints as long as all Z* extensions are grouped together?

> 
> On Sat, Apr 23, 2022 at 10:26 PM Tsukasa OI via Gcc-patches
> <gcc-patches@gcc.gnu.org> wrote:
>>
>> This commit fixes canonical extension order to follow the RISC-V ISA
>> Manual draft-20210402-1271737 or later.
>>
>> gcc/ChangeLog:
>>
>>         * common/config/riscv/riscv-common.cc (riscv_supported_std_ext):
>>         Fix "K" extension prefix to be placed before "J".
>> ---
>>  gcc/common/config/riscv/riscv-common.cc | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/gcc/common/config/riscv/riscv-common.cc b/gcc/common/config/riscv/riscv-common.cc
>> index 1501242e296..0b0ec2c4ec5 100644
>> --- a/gcc/common/config/riscv/riscv-common.cc
>> +++ b/gcc/common/config/riscv/riscv-common.cc
>> @@ -594,7 +594,7 @@ riscv_subset_list::lookup (const char *subset, int major_version,
>>  static const char *
>>  riscv_supported_std_ext (void)
>>  {
>> -  return "mafdqlcbjktpvn";
>> +  return "mafdqlcbkjtpvn";
>>  }
>>
>>  /* Parsing subset version.
>> --
>> 2.32.0
>>
> 

Re: [PATCH 1/1] RISC-V: Fix canonical extension order (K and J)

[Andrew Waterman]

Neither K nor J is an extension that exists, and so it doesn't make
sense to mandate any particular ordering.  The better change would be
to delete the letters `k' and `j' from that string, so that we aren't
enforcing constraints that don't serve a useful purpose.

cf. https://github.com/riscv/riscv-isa-manual/commit/f5f9c27010b69a015958ffebe1ac5a34f8776dff

On Sat, Apr 23, 2022 at 10:26 PM Tsukasa OI via Gcc-patches
<gcc-patches@gcc.gnu.org> wrote:
>
> This commit fixes canonical extension order to follow the RISC-V ISA
> Manual draft-20210402-1271737 or later.
>
> gcc/ChangeLog:
>
>         * common/config/riscv/riscv-common.cc (riscv_supported_std_ext):
>         Fix "K" extension prefix to be placed before "J".
> ---
>  gcc/common/config/riscv/riscv-common.cc | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/gcc/common/config/riscv/riscv-common.cc b/gcc/common/config/riscv/riscv-common.cc
> index 1501242e296..0b0ec2c4ec5 100644
> --- a/gcc/common/config/riscv/riscv-common.cc
> +++ b/gcc/common/config/riscv/riscv-common.cc
> @@ -594,7 +594,7 @@ riscv_subset_list::lookup (const char *subset, int major_version,
>  static const char *
>  riscv_supported_std_ext (void)
>  {
> -  return "mafdqlcbjktpvn";
> +  return "mafdqlcbkjtpvn";
>  }
>
>  /* Parsing subset version.
> --
> 2.32.0
>

[PATCH 1/1] RISC-V: Fix canonical extension order (K and J)

[Tsukasa OI]

This commit fixes canonical extension order to follow the RISC-V ISA
Manual draft-20210402-1271737 or later.

gcc/ChangeLog:

	* common/config/riscv/riscv-common.cc (riscv_supported_std_ext):
	Fix "K" extension prefix to be placed before "J".
---
 gcc/common/config/riscv/riscv-common.cc | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/gcc/common/config/riscv/riscv-common.cc b/gcc/common/config/riscv/riscv-common.cc
index 1501242e296..0b0ec2c4ec5 100644
--- a/gcc/common/config/riscv/riscv-common.cc
+++ b/gcc/common/config/riscv/riscv-common.cc
@@ -594,7 +594,7 @@ riscv_subset_list::lookup (const char *subset, int major_version,
 static const char *
 riscv_supported_std_ext (void)
 {
-  return "mafdqlcbjktpvn";
+  return "mafdqlcbkjtpvn";
 }
 
 /* Parsing subset version.
-- 
2.32.0