Skip to contentSkip to conversation

x86: Remove STACKPROTECTOR_ALL

DISCUSSION Thomas Backlund 2 participants 3 messages Nov 10, 2009 – Dec 6, 2009 1 message not in the archive

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

#1 lore raw

I think this one should go to 2.6.31.x too ...

Quoted 36 lines
Gitweb:     http://git.kernel.org/linus/14a3f40aafacde1dfd6912327ae14df4baf10304
Commit:     14a3f40aafacde1dfd6912327ae14df4baf10304
Parent:     02dd0a0613e0d84c7dd8315e3fe6204d005b7c79
Author:     Arjan van de Ven <[email protected]>
AuthorDate: Fri Oct 23 07:31:01 2009 -0700
Committer:  Ingo Molnar <[email protected]>
CommitDate: Fri Oct 23 16:35:23 2009 +0200

x86: Remove STACKPROTECTOR_ALL

STACKPROTECTOR_ALL has a really high overhead (runtime and stack footprint) and is not really worth it protection wise (the normal STACKPROTECTOR is in effect for all functions with buffers already), so lets just remove the option entirely.

LKML-Reference: <[email protected]>

---
 arch/x86/Kconfig  |    4 ----
 arch/x86/Makefile |    1 -
 2 files changed, 0 insertions(+), 5 deletions(-)
arch/x86/Kconfig +0 -2
index 07e0114..72ace95 100644
@@ -1443,12 +1443,8 @@ config SECCOMP
 
 	  If unsure, say Y. Only embedded should say N here.
 
-config CC_STACKPROTECTOR_ALL
-	bool
#2 lore raw
Quoted 23 lines · Thomas Backlund

I think this one should go to 2.6.31.x too ...

Quoted 21 lines
Gitweb:     http://git.kernel.org/linus/14a3f40aafacde1dfd6912327ae14df4baf10304
Commit:     14a3f40aafacde1dfd6912327ae14df4baf10304
Parent:     02dd0a0613e0d84c7dd8315e3fe6204d005b7c79
Author:     Arjan van de Ven <[email protected]>
AuthorDate: Fri Oct 23 07:31:01 2009 -0700
Committer:  Ingo Molnar <[email protected]>
CommitDate: Fri Oct 23 16:35:23 2009 +0200

x86: Remove STACKPROTECTOR_ALL

STACKPROTECTOR_ALL has a really high overhead (runtime and stack footprint) and is not really worth it protection wise (the normal STACKPROTECTOR is in effect for all functions with buffers already), so lets just remove the option entirely.

LKML-Reference: <[email protected]>

It doesn't really "fix" anything, so I'd prefer not too.

thanks,

greg k-h

#3 lore raw

Greg KH skrev:

Quoted 18 lines · Thomas Backlund
Quoted 17 lines

I think this one should go to 2.6.31.x too ...

Quoted 15 lines
Gitweb:     http://git.kernel.org/linus/14a3f40aafacde1dfd6912327ae14df4baf10304
Commit:     14a3f40aafacde1dfd6912327ae14df4baf10304
Parent:     02dd0a0613e0d84c7dd8315e3fe6204d005b7c79
Author:     Arjan van de Ven <[email protected]>
AuthorDate: Fri Oct 23 07:31:01 2009 -0700
Committer:  Ingo Molnar <[email protected]>
CommitDate: Fri Oct 23 16:35:23 2009 +0200

x86: Remove STACKPROTECTOR_ALL

STACKPROTECTOR_ALL has a really high overhead (runtime and stack footprint) and is not really worth it protection wise (the normal STACKPROTECTOR is in effect for all functions with buffers already), so lets just remove the option entirely.

Quoted 9 lines
Quoted 6 lines
Quoted 6 lines

LKML-Reference: <[email protected]>

It doesn't really "fix" anything, so I'd prefer not too.

The reason for I suggested it was that is's supposed to remove some bloating, and reportedly xfs from blowing up:

http://marc.info/?l=linux-kernel&m=125614028227106&w=2

But anyway, it's your call...

--
Thomas