From: Richard Sandiford <richard.sandiford@arm.com>
To: "Kewen.Lin" <linkw@linux.ibm.com>
Cc: Richard Biener <richard.guenther@gmail.com>,
GCC Patches <gcc-patches@gcc.gnu.org>,
Segher Boessenkool <segher@kernel.crashing.org>,
Peter Bergner <bergner@linux.ibm.com>
Subject: Re: [PATCH] vect: Fold LEN_{LOAD,STORE} if it's for the whole vector [PR107412]
Date: Thu, 01 Dec 2022 11:28:00 +0000 [thread overview]
Message-ID: <mpto7sn4b3z.fsf@arm.com> (raw)
In-Reply-To: <20f398c0-4eb9-908d-d782-445b9ce5a79d@linux.ibm.com> (Kewen Lin's message of "Mon, 28 Nov 2022 10:57:29 +0800")
"Kewen.Lin" <linkw@linux.ibm.com> writes:
> Hi Richard,
>
> on 2022/11/24 17:24, Richard Sandiford wrote:
>> "Kewen.Lin" <linkw@linux.ibm.com> writes:
>>> Hi,
>>>
>>> As the test case in PR107412 shows, we can fold IFN .LEN_{LOAD,
>>> STORE} into normal vector load/store if the given length is known
>>> to be equal to the length of the whole vector. It would help to
>>> improve overall cycles as normally the latency of vector access
>>> with length in bytes is bigger than normal vector access, and it
>>> also saves the preparation for length if constant length can not
>>> be encoded into instruction (such as on power).
>>>
>>> Bootstrapped and regtested on x86_64-redhat-linux,
>>> aarch64-linux-gnu and powerpc64{,le}-linux-gnu.
>>>
>>> Is it ok for trunk?
>>>
>>> BR,
>>> Kewen
>>> -----
>>> PR tree-optimization/107412
>>>
>>> gcc/ChangeLog:
>>>
>>> * gimple-fold.cc (gimple_fold_mask_load_store_mem_ref): Rename to ...
>>> (gimple_fold_partial_load_store_mem_ref): ... this, add one parameter
>>> mask_p indicating it's for mask or length, and add some handlings for
>>> IFN LEN_{LOAD,STORE}.
>>> (gimple_fold_mask_load): Rename to ...
>>> (gimple_fold_partial_load): ... this, add one parameter mask_p.
>>> (gimple_fold_mask_store): Rename to ...
>>> (gimple_fold_partial_store): ... this, add one parameter mask_p.
>>> (gimple_fold_call): Add the handlings for IFN LEN_{LOAD,STORE},
>>> and adjust calls on gimple_fold_mask_load_store_mem_ref to
>>> gimple_fold_partial_load_store_mem_ref.
>>
>> Sorry to reply to late (still catching up on email), but:
>>
>>>
>>> gcc/testsuite/ChangeLog:
>>>
>>> * gcc.target/powerpc/pr107412.c: New test.
>>> * gcc.target/powerpc/p9-vec-length-epil-8.c: Adjust scan times for
>>> folded LEN_LOAD.
>>> ---
>>> gcc/gimple-fold.cc | 57 ++++++++++++++-----
>>> .../gcc.target/powerpc/p9-vec-length-epil-8.c | 2 +-
>>> gcc/testsuite/gcc.target/powerpc/pr107412.c | 19 +++++++
>>> 3 files changed, 64 insertions(+), 14 deletions(-)
>>> create mode 100644 gcc/testsuite/gcc.target/powerpc/pr107412.c
>>>
>>> diff --git a/gcc/gimple-fold.cc b/gcc/gimple-fold.cc
>>> index a1704784bc9..e3a087defa6 100644
>>> --- a/gcc/gimple-fold.cc
>>> +++ b/gcc/gimple-fold.cc
>>> @@ -5370,19 +5370,39 @@ arith_overflowed_p (enum tree_code code, const_tree type,
>>> return wi::min_precision (wres, sign) > TYPE_PRECISION (type);
>>> }
>>>
>>> -/* If IFN_MASK_LOAD/STORE call CALL is unconditional, return a MEM_REF
>>> +/* If IFN_{MASK,LEN}_LOAD/STORE call CALL is unconditional, return a MEM_REF
>>> for the memory it references, otherwise return null. VECTYPE is the
>>> - type of the memory vector. */
>>> + type of the memory vector. MASK_P indicates it's for MASK if true,
>>> + otherwise it's for LEN. */
>>>
>>> static tree
>>> -gimple_fold_mask_load_store_mem_ref (gcall *call, tree vectype)
>>> +gimple_fold_partial_load_store_mem_ref (gcall *call, tree vectype, bool mask_p)
>>> {
>>> tree ptr = gimple_call_arg (call, 0);
>>> tree alias_align = gimple_call_arg (call, 1);
>>> - tree mask = gimple_call_arg (call, 2);
>>> - if (!tree_fits_uhwi_p (alias_align) || !integer_all_onesp (mask))
>>> + if (!tree_fits_uhwi_p (alias_align))
>>> return NULL_TREE;
>>>
>>> + if (mask_p)
>>> + {
>>> + tree mask = gimple_call_arg (call, 2);
>>> + if (!integer_all_onesp (mask))
>>> + return NULL_TREE;
>>> + } else {
>>
>> Minor nit: }, else, and { should be on separate lines. But the thing
>> I actually wanted to say was...
>
> Thanks for catching, I must have forgotten to reformat these lines.
>
>>
>>> + tree basic_len = gimple_call_arg (call, 2);
>>> + if (!tree_fits_uhwi_p (basic_len))
>>> + return NULL_TREE;
>>> + unsigned int nargs = gimple_call_num_args (call);
>>> + tree bias = gimple_call_arg (call, nargs - 1);
>>> + gcc_assert (tree_fits_uhwi_p (bias));
>>> + tree biased_len = int_const_binop (MINUS_EXPR, basic_len, bias);
>>> + unsigned int len = tree_to_uhwi (biased_len);
>>> + unsigned int vect_len
>>> + = GET_MODE_SIZE (TYPE_MODE (vectype)).to_constant ();
>>> + if (vect_len != len)
>>> + return NULL_TREE;
>>
>> Using "unsigned int" truncates the value. I realise that's probably
>> safe in this context, since large values have undefined behaviour.
>> But it still seems better to use an untruncated type, so that it
>> looks less like an oversight. (I think this is one case where "auto"
>> can help, since it gets the type right automatically.)
>>
>> It would also be better to avoid the to_constant, since we haven't
>> proven is_constant. How about:
>>
>> tree basic_len = gimple_call_arg (call, 2);
>> if (!poly_int_tree_p (basic_len))
>> return NULL_TREE;
>> unsigned int nargs = gimple_call_num_args (call);
>> tree bias = gimple_call_arg (call, nargs - 1);
>> gcc_assert (TREE_CODE (bias) == INTEGER_CST);
>> if (maybe_ne (wi::to_poly_widest (basic_len) - wi::to_widest (bias),
>> GET_MODE_SIZE (TYPE_MODE (vectype))))
>> return NULL_TREE;
>>
>> which also avoids using tree arithmetic for the subtraction?
>
> I agree your proposed code has better robustness, thanks!
>
> Sorry that the original patch was committed, I made a patch as attached.
> It's bootstrapped and regresss-tested on powerpc64-linux-gnu P8, and
> powerpc64le-linux-gnu P9 and P10.
>
> Is it ok for trunk?
OK, thanks.
Richard
> BR,
> Kewen
>
> From 3984a7f86a35d13e1fd40bc0c12ed5ad5b234047 Mon Sep 17 00:00:00 2001
> From: Kewen Lin <linkw@linux.ibm.com>
> Date: Sun, 27 Nov 2022 20:29:57 -0600
> Subject: [PATCH] gimple-fold: Refine gimple_fold_partial_load_store_mem_ref
>
> Following Richard's review comments, this patch is to use
> untruncated type for the length used for IFN_LEN_{LOAD,STORE}
> instead of "unsigned int" for better robustness. It also
> avoid to use to_constant and tree arithmetic for subtraction.
>
> Co-authored-by: Richard Sandiford <richard.sandiford@arm.com>
>
> gcc/ChangeLog:
>
> * gimple-fold.cc (gimple_fold_partial_load_store_mem_ref): Use
> untruncated type for the length, and avoid to_constant and tree
> arithmetic for subtraction.
> ---
> gcc/gimple-fold.cc | 15 +++++++--------
> 1 file changed, 7 insertions(+), 8 deletions(-)
>
> diff --git a/gcc/gimple-fold.cc b/gcc/gimple-fold.cc
> index c2d9c806aee..88d14c7adcc 100644
> --- a/gcc/gimple-fold.cc
> +++ b/gcc/gimple-fold.cc
> @@ -5387,18 +5387,17 @@ gimple_fold_partial_load_store_mem_ref (gcall *call, tree vectype, bool mask_p)
> tree mask = gimple_call_arg (call, 2);
> if (!integer_all_onesp (mask))
> return NULL_TREE;
> - } else {
> + }
> + else
> + {
> tree basic_len = gimple_call_arg (call, 2);
> - if (!tree_fits_uhwi_p (basic_len))
> + if (!poly_int_tree_p (basic_len))
> return NULL_TREE;
> unsigned int nargs = gimple_call_num_args (call);
> tree bias = gimple_call_arg (call, nargs - 1);
> - gcc_assert (tree_fits_shwi_p (bias));
> - tree biased_len = int_const_binop (MINUS_EXPR, basic_len, bias);
> - unsigned int len = tree_to_uhwi (biased_len);
> - unsigned int vect_len
> - = GET_MODE_SIZE (TYPE_MODE (vectype)).to_constant ();
> - if (vect_len != len)
> + gcc_assert (TREE_CODE (bias) == INTEGER_CST);
> + if (maybe_ne (wi::to_poly_widest (basic_len) - wi::to_widest (bias),
> + GET_MODE_SIZE (TYPE_MODE (vectype))))
> return NULL_TREE;
> }
prev parent reply other threads:[~2022-12-01 11:28 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-11-02 7:59 Kewen.Lin
2022-11-05 11:40 ` Richard Biener
2022-11-24 9:24 ` Richard Sandiford
2022-11-28 2:57 ` Kewen.Lin
2022-12-01 11:28 ` Richard Sandiford [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=mpto7sn4b3z.fsf@arm.com \
--to=richard.sandiford@arm.com \
--cc=bergner@linux.ibm.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=linkw@linux.ibm.com \
--cc=richard.guenther@gmail.com \
--cc=segher@kernel.crashing.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).