public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
To: "H. Peter Anvin" <hpa@zytor.com>
Cc: akpm@linux-foundation.org, Ingo Molnar <mingo@elte.hu>,
	linux-kernel@vger.kernel.org, Andi Kleen <andi@firstfloor.org>,
	Rusty Russell <rusty@rustcorp.com.au>, Andi Kleen <ak@muc.de>,
	Chuck Ebbert <cebbert@redhat.com>,
	Christoph Hellwig <hch@infradead.org>,
	Jeremy Fitzhardinge <jeremy@goop.org>,
	Thomas Gleixner <tglx@linutronix.de>,
	Ingo Molnar <mingo@redhat.com>, Adrian Bunk <bunk@stusta.de>,
	Alexey Dobriyan <adobriyan@gmail.com>,
	akpm@osdl.org
Subject: Re: [patch 13/17] Immediate Values - x86 Optimization (updated)
Date: Wed, 9 Apr 2008 19:15:34 -0400	[thread overview]
Message-ID: <20080409231534.GA7842@Krystal> (raw)
In-Reply-To: <47FD4424.20907@zytor.com>

* H. Peter Anvin (hpa@zytor.com) wrote:
> Mathieu Desnoyers wrote:
>> Hrm, since even the nmi-safe version supports REX-prefixed instructions,
>> there is no need for an =q constraint on single-byte immediate values
>> anymore. (thanks to your "discard" section used in the nmi-safe version)
>
> WRONG!
>
> Using =r for single-byte values is incorrect for 32-bit code -- that would 
> permit %spl, %bpl, %sil, %dil which are illegal in 32-bit mode. That is not 
> the incorrect bit, it's the description that is confused.
>
> 	-hpa

Ah, right. I remembered there was something that made us use =q, but
could not remember what. I'll describe it correctly.

Therefore, this updated patch is bogus. The original one was ok, given
that we change the header to :


Immediate Values - x86 Optimization

x86 optimization of the immediate values which uses a movl with code patching
to set/unset the value used to populate the register used as variable source.

Note : a movb needs to get its value froma =q constraint.

Quoting "H. Peter Anvin" <hpa@zytor.com>

Using =r for single-byte values is incorrect for 32-bit code -- that would  
permit %spl, %bpl, %sil, %dil which are illegal in 32-bit mode. 

Changelog:
- Use text_poke_early with cr0 WP save/restore to patch the bypass. We are doing
  non atomic writes to a code region only touched by us (nobody can execute it
  since we are protected by the imv_mutex).
- Put imv_set and _imv_set in the architecture independent header.
- Use $0 instead of %2 with (0) operand.
- Add x86_64 support, ready for i386+x86_64 -> x86 merge.
- Use asm-x86/asm.h.
- Bugfix : 8 bytes 64 bits immediate value was declared as "4 bytes" in the
  immediate structure.
- Change the immediate.c update code to support variable length opcodes.
- Vastly simplified, using a busy looping IPI with interrupts disabled.
  Does not protect against NMI nor MCE.
- Pack the __imv section. Use smallest types required for size (char).
- Use imv_* instead of immediate_*.


Mathieu

-- 
Mathieu Desnoyers
Computer Engineering Ph.D. Student, Ecole Polytechnique de Montreal
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68

  reply	other threads:[~2008-04-09 23:20 UTC|newest]

Thread overview: 48+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-04-09 15:08 [patch 00/17] Text Edit Lock and Immediate Values for 2.6.25-rc8-mm1 Mathieu Desnoyers
2008-04-09 15:08 ` [patch 01/17] Kprobes - use a mutex to protect the instruction pages list Mathieu Desnoyers
2008-04-09 20:08   ` Masami Hiramatsu
2008-04-09 15:08 ` [patch 02/17] Kprobes - do not use kprobes mutex in arch code Mathieu Desnoyers
2008-04-09 20:08   ` Masami Hiramatsu
2008-04-09 15:08 ` [patch 03/17] Kprobes - declare kprobe_mutex static Mathieu Desnoyers
2008-04-09 20:08   ` Masami Hiramatsu
2008-04-09 15:08 ` [patch 04/17] x86 - Enhance DEBUG_RODATA support - alternatives Mathieu Desnoyers
2008-04-09 15:08 ` [patch 05/17] x86 Fix text_poke for vmalloced pages Mathieu Desnoyers
2008-04-09 15:08 ` [patch 06/17] x86 - Enhance DEBUG_RODATA support for hotplug and kprobes Mathieu Desnoyers
2008-04-09 15:08 ` [patch 07/17] Text Edit Lock - Architecture Independent Code Mathieu Desnoyers
2008-04-09 15:08 ` [patch 08/17] Text Edit Lock - kprobes architecture independent support Mathieu Desnoyers
2008-04-09 15:08 ` [patch 09/17] Add all cpus option to stop machine run Mathieu Desnoyers
2008-04-09 18:10   ` Alexey Dobriyan
2008-04-09 18:24     ` Andi Kleen
2008-04-10  3:34       ` Rusty Russell
2008-04-10  4:26       ` KOSAKI Motohiro
2008-04-09 18:54     ` Mathieu Desnoyers
2008-04-09 15:08 ` [patch 10/17] Immediate Values - Architecture Independent Code Mathieu Desnoyers
2008-04-09 15:08 ` [patch 11/17] Implement immediate update via stop_machine_run Mathieu Desnoyers
2008-04-10  8:04   ` KOSAKI Motohiro
2008-04-10 20:01     ` Mathieu Desnoyers
2008-04-11  4:50       ` KOSAKI Motohiro
2008-04-09 15:08 ` [patch 12/17] Immediate Values - Kconfig menu in EMBEDDED Mathieu Desnoyers
2008-04-10  3:23   ` Rusty Russell
2008-04-10 19:32     ` [patch 12/17] Immediate Values - Kconfig menu in EMBEDDED (updated) Mathieu Desnoyers
2008-04-10 21:54       ` Rusty Russell
2008-04-14 23:52         ` Mathieu Desnoyers
2008-04-09 15:08 ` [patch 13/17] Immediate Values - x86 Optimization Mathieu Desnoyers
2008-04-09 18:01   ` H. Peter Anvin
2008-04-09 19:08     ` Mathieu Desnoyers
2008-04-09 22:33       ` H. Peter Anvin
2008-04-10  0:42         ` Mathieu Desnoyers
2008-04-10  0:47           ` H. Peter Anvin
2008-04-09 20:21     ` [patch 13/17] Immediate Values - x86 Optimization (updated) Mathieu Desnoyers
2008-04-09 22:33       ` H. Peter Anvin
2008-04-09 23:15         ` Mathieu Desnoyers [this message]
2008-04-09 15:08 ` [patch 14/17] Add text_poke and sync_core to powerpc Mathieu Desnoyers
2008-04-09 15:08 ` [patch 15/17] Immediate Values - Powerpc Optimization Mathieu Desnoyers
2008-04-09 15:08 ` [patch 16/17] Immediate Values - Documentation Mathieu Desnoyers
2008-04-10  3:33   ` Rusty Russell
2008-04-11  1:16     ` Mathieu Desnoyers
2008-04-11 15:06       ` Rusty Russell
2008-04-15  0:12         ` Mathieu Desnoyers
2008-04-11 13:44     ` [RFC PATCH] Immediate Values Support init Mathieu Desnoyers
2008-04-09 15:08 ` [patch 17/17] Scheduler Profiling - Use Immediate Values Mathieu Desnoyers
2008-04-10  4:23 ` [patch 00/17] Text Edit Lock and Immediate Values for 2.6.25-rc8-mm1 KOSAKI Motohiro
2008-04-10  7:31 ` Takashi Nishiie

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=20080409231534.GA7842@Krystal \
    --to=mathieu.desnoyers@polymtl.ca \
    --cc=adobriyan@gmail.com \
    --cc=ak@muc.de \
    --cc=akpm@linux-foundation.org \
    --cc=akpm@osdl.org \
    --cc=andi@firstfloor.org \
    --cc=bunk@stusta.de \
    --cc=cebbert@redhat.com \
    --cc=hch@infradead.org \
    --cc=hpa@zytor.com \
    --cc=jeremy@goop.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=mingo@redhat.com \
    --cc=rusty@rustcorp.com.au \
    --cc=tglx@linutronix.de \
    /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