login
Header Space

 
 

Re: [git pull] x86 PAT changes

Score:
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
To: Linus Torvalds <torvalds@...>
Cc: H. Peter Anvin <hpa@...>, <linux-kernel@...>, Andrew Morton <akpm@...>, Thomas Gleixner <tglx@...>
Date: Saturday, April 26, 2008 - 4:56 am

* Linus Torvalds <torvalds@linux-foundation.org> wrote:


You are right, there should be no architectural need to make PAT 
dependent on nonpromisc-devmem, and thus the patch below should be safe.

In theory even mmap() of /dev/mem should be safe this way - as all 
memtypes are properly tracked.

The thinking behind this dependency was three-fold:

- historic: from the days when the PAT patchset didnt do fully correct 
  tracking yet

- practical: that PAT would be utilized in newer distros on newer 
  systems - with older distros on older systems not really wanting 
  (or needing) neither /dev/mem restrictions nor PAT

- paranoia: one less degree of freedom to take into account

	Ingo

----------------------->
Subject: x86 PAT: decouple from nonpromisc devmem
From: Ingo Molnar <mingo@elte.hu>
Date: Sat Apr 26 10:26:52 CEST 2008

Linus pointed it out that PAT should not depend on NONPROMISC_DEVMEM.

Also make PAT non-default.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 arch/x86/Kconfig |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

Index: linux-x86.q/arch/x86/Kconfig
===================================================================
--- linux-x86.q.orig/arch/x86/Kconfig
+++ linux-x86.q/arch/x86/Kconfig
@@ -1042,9 +1042,9 @@ config MTRR
 	  See <file:Documentation/mtrr.txt> for more information.
 
 config X86_PAT
-	def_bool y
+	bool
 	prompt "x86 PAT support"
-	depends on MTRR && NONPROMISC_DEVMEM
+	depends on MTRR
 	help
 	  Use PAT attributes to setup page level cache control.
 
--
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]

Messages in current thread:
[git pull] x86 PAT changes, Ingo Molnar, (Thu Apr 24, 6:56 pm)
Re: [git pull] x86 PAT changes, Gabriel C, (Sat Apr 26, 9:40 am)
Re: [git pull] x86 PAT changes, Ingo Molnar, (Sat Apr 26, 10:42 am)
Re: [git pull] x86 PAT changes, Gabriel C, (Sat Apr 26, 11:37 am)
Re: [git pull] x86 PAT changes, Ingo Molnar, (Sat Apr 26, 11:41 am)
Re: [git pull] x86 PAT changes, Linus Torvalds, (Sat Apr 26, 12:43 pm)
Re: [git pull] x86 PAT changes, Linus Torvalds, (Fri Apr 25, 7:43 pm)
Re: [git pull] x86 PAT changes, Ingo Molnar, (Sat Apr 26, 5:57 am)
Re: [git pull] x86 PAT changes, H. Peter Anvin, (Fri Apr 25, 8:06 pm)
Re: [git pull] x86 PAT changes, Linus Torvalds, (Fri Apr 25, 9:12 pm)
Re: [git pull] x86 PAT changes, Ingo Molnar, (Sat Apr 26, 4:56 am)
RE: [git pull] x86 PAT changes, Pallipadi, Venkatesh, (Sat Apr 26, 12:54 pm)
RE: [git pull] x86 PAT changes, Linus Torvalds, (Sat Apr 26, 1:15 pm)
Re: [git pull] x86 PAT changes, Venki Pallipadi, (Sat Apr 26, 2:32 pm)
[patch] x86, PAT: disable /dev/mem mmap RAM with PAT, Ingo Molnar, (Sat Apr 26, 3:07 pm)
speck-geostationary