All the mail mirrored from lore.kernel.org
 help / color / mirror / Atom feed
From: Dave.Martin@arm.com (Dave Martin)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH v2 00/10] arm64: Use BRK instruction for generic BUG traps
Date: Mon, 13 Jul 2015 14:25:47 +0100	[thread overview]
Message-ID: <1436793967-7138-1-git-send-email-Dave.Martin@arm.com> (raw)

Changes since v1:

 * Modified BRK immediate for BUG so that it doesn't overlap the range
   allocated for KGDB.

 * Typo fixes.

 * Don't assume that BUG() kills the thread, so that catching BUGs in
   kgdb works again.

 * Separate header for the BRK immediates removed, at Will's request.
   I've retained the other refactoring since it contains useful tidy-
   ups, but some of that could go away if desired.


Original cover letter:

Currently, the minimal default BUG() implementation from asm-generic is
used for arm64.

This series uses the BRK software breakpoint instruction to generate a
trap instead, similarly to most other arches, with the generic BUG code
generating the dmesg boilerplate.  This eliminates a fair amount of
inlined code at BUG() and WARN() sites.


This work makes it look increasingly desirable to collect BRK immediates
together in one place.  Patches 1-7 do some refactoring to prepare for
this, and patch 8 moves the definitions to a fresh header, <asm/brk.h>.

Patch 9 provides the BRK-based GENERIC_BUG support for arm64.

A side-effect of this change is that WARNs are now generated via a
different bit of generic code (lib/bug.c:report_bug()) that no longer
prints a backtrace (compare kernel/panic.c:warn_slowpath_common()).)  I
will post a separate mini-RFC series to address that in the generic
code.  Patch 10 hacks a backtrace back into the arm64 arch code in the
meantime.


Comments and testing welcome.

Quick testing suggests a kernel size reduction of ~110K for arm64
defconfig.


Dave Martin (10):
  arm64/debug: Eliminate magic number for size of BRK instruction
  arm64/debug: Mask off all reserved bits from generated ESR values
  arm64: esr.h type fixes and cleanup
  arm64/debug: Eliminate magic number from ESR template definition
  arm64/debug: More consistent naming for the BRK ESR template macro
  arm64/debug: Move BRK ESR template macro into <asm/esr.h>
  arm64/debug: Simplify BRK insn opcode declarations
  arm64/debug: Add missing #include
  arm64/BUG: Use BRK instruction for generic BUG traps
  arm64/BUG: Show explicit backtrace for WARNs

 arch/arm64/Kconfig                      |    8 ++
 arch/arm64/include/asm/bug.h            |   64 +++++++++++++++
 arch/arm64/include/asm/debug-monitors.h |   36 +++------
 arch/arm64/include/asm/esr.h            |  135 ++++++++++++++++---------------
 arch/arm64/include/asm/memory.h         |    3 +-
 arch/arm64/kernel/kgdb.c                |   12 +--
 arch/arm64/kernel/traps.c               |   61 +++++++++++++-
 arch/arm64/mm/fault.c                   |   12 ++-
 8 files changed, 233 insertions(+), 98 deletions(-)
 create mode 100644 arch/arm64/include/asm/bug.h

-- 
1.7.10.4

             reply	other threads:[~2015-07-13 13:25 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-07-13 13:25 Dave Martin [this message]
2015-07-13 13:25 ` [PATCH v2 01/10] arm64/debug: Eliminate magic number for size of BRK instruction Dave Martin
2015-07-13 14:47   ` Mark Rutland
2015-07-13 13:25 ` [PATCH v2 02/10] arm64/debug: Mask off all reserved bits from generated ESR values Dave Martin
2015-07-13 14:14   ` Mark Rutland
2015-07-13 14:22     ` Dave Martin
2015-07-13 13:25 ` [PATCH v2 03/10] arm64: esr.h type fixes and cleanup Dave Martin
2015-07-14 15:54   ` Mark Rutland
2015-07-14 16:53     ` Dave Martin
2015-07-13 13:25 ` [PATCH v2 04/10] arm64/debug: Eliminate magic number from ESR template definition Dave Martin
2015-07-13 14:16   ` Mark Rutland
2015-07-13 13:25 ` [PATCH v2 05/10] arm64/debug: More consistent naming for the BRK ESR template macro Dave Martin
2015-07-13 14:19   ` Mark Rutland
2015-07-13 13:25 ` [PATCH v2 06/10] arm64/debug: Move BRK ESR template macro into <asm/esr.h> Dave Martin
2015-07-13 14:18   ` Mark Rutland
2015-07-13 13:25 ` [PATCH v2 07/10] arm64/debug: Simplify BRK insn opcode declarations Dave Martin
2015-07-13 14:38   ` Mark Rutland
2015-07-13 13:25 ` [PATCH v2 08/10] arm64/debug: Add missing #include Dave Martin
2015-07-13 14:34   ` Mark Rutland
2015-07-13 14:44     ` Dave Martin
2015-07-13 14:45       ` Mark Rutland
2015-07-13 13:25 ` [PATCH v2 09/10] arm64/BUG: Use BRK instruction for generic BUG traps Dave Martin
2015-07-13 16:43   ` Mark Rutland
2015-07-13 16:51     ` Dave Martin
2015-07-13 16:56       ` Mark Rutland
2015-07-13 17:05         ` Dave P Martin
2015-07-14 10:20         ` Dave Martin
2015-07-14 11:09           ` Mark Rutland
2015-07-14 11:34             ` Dave Martin
2015-07-14 15:51               ` Mark Rutland
2015-07-14 16:53                 ` Dave Martin
2015-07-14 16:11   ` Catalin Marinas
2015-07-14 16:55     ` Dave Martin
2015-07-13 13:25 ` [PATCH v2 10/10] arm64/BUG: Show explicit backtrace for WARNs Dave Martin

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=1436793967-7138-1-git-send-email-Dave.Martin@arm.com \
    --to=dave.martin@arm.com \
    --cc=linux-arm-kernel@lists.infradead.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 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.