public inbox for gcc-bugs@sourceware.org help / color / mirror / Atom feed
From: "jakub at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org> To: gcc-bugs@gcc.gnu.org Subject: [Bug c/102989] Implement C2x's n2763 (_BitInt) Date: Mon, 26 Jun 2023 18:48:34 +0000 [thread overview] Message-ID: <bug-102989-4-HuMFR35REn@http.gcc.gnu.org/bugzilla/> (raw) In-Reply-To: <bug-102989-4@http.gcc.gnu.org/bugzilla/> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102989 --- Comment #70 from Jakub Jelinek <jakub at gcc dot gnu.org> --- For right shifts, I wonder if we shouldn't emit inline (perhaps with exception of -Os) something like: __attribute__((noipa)) void ashiftrt575 (unsigned long *p, unsigned long *q, int n) { int prec = 575; int n1 = n & 63; int n2 = n / 64; int n3 = n1 != 0; int n4 = (-n1) & 63; unsigned long ext; int i; for (i = n2; i < prec / 64 - n3; ++i) p[i - n2] = (q[i] >> n1) | (q[i + n3] << n4); ext = ((signed long) (q[prec / 64] << (64 - (prec & 63)))) >> (64 - (prec & 63)); if (n1 && i == prec / 64 - n3) { p[i - n2] = (q[i] >> n1) | (ext << n4); ++i; } i -= n2; p[i] = ((signed long) ext) >> n1; ext = ((signed long) ext) >> 63; for (++i; i < prec / 64 + 1; ++i) p[i] = ext; } __attribute__((noipa)) void lshiftrt575 (unsigned long *p, unsigned long *q, int n) { int prec = 575; int n1 = n & 63; int n2 = n / 64; int n3 = n1 != 0; int n4 = (-n1) & 63; unsigned long ext; int i; for (i = n2; i < prec / 64 - n3; ++i) p[i - n2] = (q[i] >> n1) | (q[i + n3] << n4); ext = q[prec / 64] & ((1UL << (prec % 64)) - 1); if (n1 && i == prec / 64 - n3) { p[i - n2] = (q[i] >> n1) | (ext << n4); ++i; } i -= n2; p[i] = ext >> n1; ext = 0; for (++i; i < prec / 64 + 1; ++i) p[i] = 0; } (for _BitInt(575) and 64-bit limb little endian). If the shift count is constant, it will allow further optimizations, and if e.g. get_nonzero_bits tells us that n is variable but multiple of limb precision, we can optimize some more as well. Looking at what LLVM does, they seem to sign extend in memory to twice as many bits and then just use an unrolled loop without any conditionals, but that doesn't look well for memory usage etc.
next prev parent reply other threads:[~2023-06-26 18:48 UTC|newest] Thread overview: 119+ messages / expand[flat|nested] mbox.gz Atom feed top 2021-10-28 17:41 [Bug c/102989] New: Add Clang's _ExtInt(N) colomar.6.4.3 at gmail dot com 2021-10-28 17:54 ` [Bug c/102989] " colomar.6.4.3 at gmail dot com 2021-10-28 17:57 ` colomar.6.4.3 at gmail dot com 2021-10-28 17:57 ` colomar.6.4.3 at gmail dot com 2021-10-28 18:01 ` jakub at gcc dot gnu.org 2021-10-28 18:11 ` colomar.6.4.3 at gmail dot com 2021-10-28 21:41 ` joseph at codesourcery dot com 2021-10-28 21:47 ` [Bug c/102989] Implement C2x's n2763 (_BitInt) pinskia at gcc dot gnu.org 2021-10-28 21:49 ` pinskia at gcc dot gnu.org 2021-11-11 19:42 ` pinskia at gcc dot gnu.org 2021-11-11 19:58 ` colomar.6.4.3 at gmail dot com 2021-11-11 21:27 ` joseph at codesourcery dot com 2022-10-25 12:14 ` jakub at gcc dot gnu.org 2022-10-25 15:25 ` hjl.tools at gmail dot com 2022-10-25 20:32 ` joseph at codesourcery dot com 2022-10-25 20:42 ` hjl.tools at gmail dot com 2022-10-25 20:45 ` jakub at gcc dot gnu.org 2022-10-25 21:05 ` segher at gcc dot gnu.org 2022-10-25 21:05 ` hjl.tools at gmail dot com 2022-10-25 21:09 ` pinskia at gcc dot gnu.org 2022-10-25 21:10 ` jakub at gcc dot gnu.org 2022-10-25 21:30 ` segher at gcc dot gnu.org 2022-10-25 21:50 ` segher at gcc dot gnu.org 2022-10-26 6:29 ` uweigand at gcc dot gnu.org 2022-10-26 6:50 ` jakub at gcc dot gnu.org 2022-10-26 8:35 ` redi at gcc dot gnu.org 2022-10-26 17:29 ` joseph at codesourcery dot com 2022-10-28 9:47 ` rguenth at gcc dot gnu.org 2022-10-28 10:32 ` jakub at gcc dot gnu.org 2022-10-28 10:51 ` rguenther at suse dot de 2022-10-28 11:02 ` colomar.6.4.3 at gmail dot com 2022-10-28 17:27 ` pinskia at gcc dot gnu.org 2022-10-28 20:31 ` joseph at codesourcery dot com 2022-10-28 20:39 ` joseph at codesourcery dot com 2023-04-09 19:59 ` leni536 at gmail dot com 2023-04-12 22:17 ` george at bott dot gg 2023-05-11 18:00 ` jakub at gcc dot gnu.org 2023-05-11 18:21 ` jakub at gcc dot gnu.org 2023-05-11 22:10 ` joseph at codesourcery dot com 2023-05-12 7:46 ` jakub at gcc dot gnu.org 2023-05-12 8:41 ` rguenth at gcc dot gnu.org 2023-05-16 16:20 ` jakub at gcc dot gnu.org 2023-05-17 7:22 ` rguenth at gcc dot gnu.org 2023-05-23 12:04 ` jakub at gcc dot gnu.org 2023-05-24 11:48 ` jakub at gcc dot gnu.org 2023-05-24 12:46 ` rguenther at suse dot de 2023-05-24 13:16 ` jakub at gcc dot gnu.org 2023-05-24 13:29 ` rguenther at suse dot de 2023-05-24 14:18 ` jakub at gcc dot gnu.org 2023-05-24 14:57 ` rguenther at suse dot de 2023-05-24 15:31 ` jakub at gcc dot gnu.org 2023-05-26 16:13 ` jakub at gcc dot gnu.org 2023-05-26 16:16 ` jakub at gcc dot gnu.org 2023-05-26 17:11 ` jakub at gcc dot gnu.org 2023-06-02 10:39 ` jakub at gcc dot gnu.org 2023-06-02 10:43 ` rguenther at suse dot de 2023-06-02 10:53 ` jakub at gcc dot gnu.org 2023-06-02 17:06 ` jakub at gcc dot gnu.org 2023-06-05 7:14 ` rguenth at gcc dot gnu.org 2023-06-05 7:34 ` jakub at gcc dot gnu.org 2023-06-05 7:43 ` rguenth at gcc dot gnu.org 2023-06-05 7:58 ` jakub at gcc dot gnu.org 2023-06-05 8:21 ` rguenther at suse dot de 2023-06-05 8:33 ` jakub at gcc dot gnu.org 2023-06-06 7:13 ` rguenth at gcc dot gnu.org 2023-06-15 11:28 ` jakub at gcc dot gnu.org 2023-06-15 18:02 ` jakub at gcc dot gnu.org 2023-06-19 17:40 ` jakub at gcc dot gnu.org 2023-06-20 20:04 ` jakub at gcc dot gnu.org 2023-06-22 19:47 ` jakub at gcc dot gnu.org 2023-06-23 17:03 ` jakub at gcc dot gnu.org 2023-06-26 18:48 ` jakub at gcc dot gnu.org [this message] 2023-06-28 17:21 ` jakub at gcc dot gnu.org 2023-06-29 17:01 ` jakub at gcc dot gnu.org 2023-06-30 19:22 ` jakub at gcc dot gnu.org 2023-07-05 17:23 ` jakub at gcc dot gnu.org 2023-07-07 14:26 ` jakub at gcc dot gnu.org 2023-07-07 17:59 ` jakub at gcc dot gnu.org 2023-07-11 11:20 ` jakub at gcc dot gnu.org 2023-07-12 16:28 ` jakub at gcc dot gnu.org 2023-07-13 18:03 ` jakub at gcc dot gnu.org 2023-07-14 11:18 ` jakub at gcc dot gnu.org 2023-07-14 11:18 ` jakub at gcc dot gnu.org 2023-07-14 17:19 ` jakub at gcc dot gnu.org 2023-07-17 10:21 ` jakub at gcc dot gnu.org 2023-07-17 18:06 ` jakub at gcc dot gnu.org 2023-07-18 11:07 ` jakub at gcc dot gnu.org 2023-07-18 15:45 ` jakub at gcc dot gnu.org 2023-07-20 15:51 ` jakub at gcc dot gnu.org 2023-07-21 17:10 ` jakub at gcc dot gnu.org 2023-07-25 14:40 ` jakub at gcc dot gnu.org 2023-07-26 13:04 ` jakub at gcc dot gnu.org 2023-07-26 17:41 ` jakub at gcc dot gnu.org 2023-07-27 15:18 ` jakub at gcc dot gnu.org 2023-08-10 7:22 ` cvs-commit at gcc dot gnu.org 2023-08-10 7:23 ` cvs-commit at gcc dot gnu.org 2023-08-10 15:30 ` cvs-commit at gcc dot gnu.org 2023-08-14 21:55 ` tmgross at umich dot edu 2023-09-06 15:57 ` cvs-commit at gcc dot gnu.org 2023-09-06 15:57 ` cvs-commit at gcc dot gnu.org 2023-09-06 15:57 ` cvs-commit at gcc dot gnu.org 2023-09-06 15:57 ` cvs-commit at gcc dot gnu.org 2023-09-06 15:57 ` cvs-commit at gcc dot gnu.org 2023-09-06 15:57 ` cvs-commit at gcc dot gnu.org 2023-09-06 15:58 ` cvs-commit at gcc dot gnu.org 2023-09-06 15:58 ` cvs-commit at gcc dot gnu.org 2023-09-06 15:58 ` cvs-commit at gcc dot gnu.org 2023-09-06 15:58 ` cvs-commit at gcc dot gnu.org 2023-09-06 15:58 ` cvs-commit at gcc dot gnu.org 2023-09-06 15:58 ` cvs-commit at gcc dot gnu.org 2023-09-06 15:58 ` cvs-commit at gcc dot gnu.org 2023-09-06 15:58 ` cvs-commit at gcc dot gnu.org 2023-09-06 15:58 ` cvs-commit at gcc dot gnu.org 2023-09-07 9:21 ` cvs-commit at gcc dot gnu.org 2023-10-12 14:07 ` cvs-commit at gcc dot gnu.org 2023-10-14 7:38 ` cvs-commit at gcc dot gnu.org 2023-10-14 22:38 ` gaius at gcc dot gnu.org 2023-11-01 8:17 ` gaius at gcc dot gnu.org 2023-11-01 9:06 ` cvs-commit at gcc dot gnu.org
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=bug-102989-4-HuMFR35REn@http.gcc.gnu.org/bugzilla/ \ --to=gcc-bugzilla@gcc.gnu.org \ --cc=gcc-bugs@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: linkBe 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).