All of lore.kernel.org
 help / color / mirror / Atom feed
From: Will Deacon <will@kernel.org>
To: linux-arm-kernel@lists.infradead.org
Cc: mark.rutland@arm.com, peterz@infradead.org,
	catalin.marinas@arm.com, ndesaulniers@google.com,
	robin.murphy@arm.com, Ard.Biesheuvel@arm.com,
	andrew.murray@arm.com, natechancellor@gmail.com,
	Will Deacon <will@kernel.org>
Subject: [PATCH v5 00/10] arm64: avoid out-of-line ll/sc atomics
Date: Thu, 29 Aug 2019 16:48:24 +0100	[thread overview]
Message-ID: <20190829154834.26547-1-will@kernel.org> (raw)

Hi all,

This is a version five of the patches previously posted by Andrew here:

  v4: http://lkml.kernel.org/r/20190828175009.15457-1-andrew.murray@arm.com

The reason I'm posting this is because I spotted an issue with the above
when queuing it for 5.4 and fixing it ended up with me spinning a few
patches on top.

The basic problem is that by implementing our atomic routines using a
static key to select between the LL/SC and LSE variant, we rely on
CONFIG_JUMP_LABEL and therefore CC_HAS_ASM_GOTO because otherwise the
static key implementation itself is implementing using atomic routines,
which leads to complete disaster.

This patch series builds on top of Andrew's patches, with the following
changes:

  * Tidying up the header files in preparation for...
  * ...making LSE depend on JUMP_LABEL
  * Support for the 'K' constraint when it looks like it works
  * Minor massaging of commit logs

This means that LSE atomics are not available for in-kernel use when
building with a version of clang without 'asm goto' support. I really
don't see a way around this, but I've been told that clang-9 should
have this support so that's at least something.

Will

Cc: Ard.Biesheuvel@arm.com
Cc: peterz@infradead.org
Cc: andrew.murray@arm.com
Cc: mark.rutland@arm.com
Cc: catalin.marinas@arm.com
Cc: robin.murphy@arm.com
Cc: ndesaulniers@google.com
Cc: natechancellor@gmail.com

--->8

Andrew Murray (5):
  jump_label: Don't warn on __exit jump entries
  arm64: Use correct ll/sc atomic constraints
  arm64: atomics: avoid out-of-line ll/sc atomics
  arm64: avoid using hard-coded registers for LSE atomics
  arm64: atomics: Remove atomic_ll_sc compilation unit

Will Deacon (5):
  arm64: lse: Remove unused 'alt_lse' assembly macro
  arm64: asm: Kill 'asm/atomic_arch.h'
  arm64: lse: Make ARM64_LSE_ATOMICS depend on JUMP_LABEL
  arm64: atomics: Undefine internal macros after use
  arm64: atomics: Use K constraint when toolchain appears to support it

 arch/arm64/Kconfig                    |   1 +
 arch/arm64/Makefile                   |   9 +-
 arch/arm64/include/asm/atomic.h       |  93 +++++++-
 arch/arm64/include/asm/atomic_ll_sc.h | 215 +++++++++---------
 arch/arm64/include/asm/atomic_lse.h   | 395 ++++++++++++----------------------
 arch/arm64/include/asm/cmpxchg.h      |  45 +++-
 arch/arm64/include/asm/lse.h          |  49 ++---
 arch/arm64/lib/Makefile               |  19 --
 arch/arm64/lib/atomic_ll_sc.c         |   3 -
 kernel/jump_label.c                   |   4 +-
 10 files changed, 413 insertions(+), 420 deletions(-)
 delete mode 100644 arch/arm64/lib/atomic_ll_sc.c

-- 
2.11.0


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

             reply	other threads:[~2019-08-29 15:48 UTC|newest]

Thread overview: 44+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-08-29 15:48 Will Deacon [this message]
2019-08-29 15:48 ` [PATCH v5 01/10] jump_label: Don't warn on __exit jump entries Will Deacon
2019-08-29 15:48 ` [PATCH v5 02/10] arm64: Use correct ll/sc atomic constraints Will Deacon
2019-08-29 15:48 ` [PATCH v5 03/10] arm64: atomics: avoid out-of-line ll/sc atomics Will Deacon
2019-09-03  6:00   ` Nathan Chancellor
2019-09-03  6:39     ` Will Deacon
2019-09-03 14:31     ` Andrew Murray
2019-09-03 14:45       ` Will Deacon
2019-09-03 15:15         ` Andrew Murray
2019-09-03 15:31           ` Andrew Murray
2019-09-03 16:37             ` Will Deacon
2019-09-03 22:04               ` Andrew Murray
2019-09-03 22:35                 ` Nick Desaulniers
     [not found]                   ` <CANW9uyuRFtNKMnSwmHWt_RebJA1ADXdZfeDHc6=yaaFH2NsyWg@mail.gmail.com>
2019-09-03 22:53                     ` Nick Desaulniers
2019-09-04 10:20                       ` Will Deacon
2019-09-04 17:28                 ` Nick Desaulniers
2019-09-05 11:25                   ` Andrew Murray
2019-09-06 19:44                     ` Nick Desaulniers
2019-08-29 15:48 ` [PATCH v5 04/10] arm64: avoid using hard-coded registers for LSE atomics Will Deacon
2019-08-29 15:48 ` [PATCH v5 05/10] arm64: atomics: Remove atomic_ll_sc compilation unit Will Deacon
2019-08-29 17:47   ` Nick Desaulniers
2019-08-29 20:07     ` Tri Vo
2019-08-29 21:54       ` Will Deacon
2019-08-29 15:48 ` [PATCH v5 06/10] arm64: lse: Remove unused 'alt_lse' assembly macro Will Deacon
2019-08-29 23:39   ` Andrew Murray
2019-08-29 15:48 ` [PATCH v5 07/10] arm64: asm: Kill 'asm/atomic_arch.h' Will Deacon
2019-08-29 23:43   ` Andrew Murray
2019-08-29 15:48 ` [PATCH v5 08/10] arm64: lse: Make ARM64_LSE_ATOMICS depend on JUMP_LABEL Will Deacon
2019-08-29 23:44   ` Andrew Murray
2019-08-29 15:48 ` [PATCH v5 09/10] arm64: atomics: Undefine internal macros after use Will Deacon
2019-08-29 23:44   ` Andrew Murray
2019-08-29 15:48 ` [PATCH v5 10/10] arm64: atomics: Use K constraint when toolchain appears to support it Will Deacon
2019-08-29 16:54   ` Will Deacon
2019-08-29 17:45     ` Nick Desaulniers
2019-08-29 21:53       ` Will Deacon
2019-08-30 20:57         ` Nick Desaulniers
2019-08-30  0:08     ` Andrew Murray
2019-08-30  7:52       ` Will Deacon
2019-08-30  9:11         ` Andrew Murray
2019-08-30 10:17           ` Will Deacon
2019-08-30 11:57             ` Andrew Murray
2019-08-30 10:40           ` Mark Rutland
2019-08-30 11:53             ` Andrew Murray
2019-08-29 23:49   ` Andrew Murray

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=20190829154834.26547-1-will@kernel.org \
    --to=will@kernel.org \
    --cc=Ard.Biesheuvel@arm.com \
    --cc=andrew.murray@arm.com \
    --cc=catalin.marinas@arm.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=mark.rutland@arm.com \
    --cc=natechancellor@gmail.com \
    --cc=ndesaulniers@google.com \
    --cc=peterz@infradead.org \
    --cc=robin.murphy@arm.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.