All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH] x86: Make sure get_user_desc() doesn't sign extend.
@ 2009-11-05 10:47 Chris Lalancette
  0 siblings, 0 replies; 3+ messages in thread
From: Chris Lalancette @ 2009-11-05 10:47 UTC (permalink / raw)
  To: torvalds
  Cc: x86, mingo, mingo, kvm, pbonzini, linux-kernel, tglx, hpa,
	Chris Lalancette

The current implementation of get_user_desc() sign extends
the return value because of integer promotion rules.  For
the most part, this doesn't matter, because the top bit of
base2 is usually 0.  If, however, that bit is 1, then the
entire value will be 0xffff... which is probably not what
the caller intended.  This patch casts the entire thing
to unsigned before returning, which generates almost the
same assembly as the current code but replaces the final
"cltq" (sign extend) with a "mov %eax %eax" (zero-extend).
This fixes booting certain guests under KVM.

(sending direct to Linus this time, since I got no response
to the previous 3 submissions)

Signed-off-by: Chris Lalancette <clalance@redhat.com>
---
 arch/x86/include/asm/desc.h |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/arch/x86/include/asm/desc.h b/arch/x86/include/asm/desc.h
index e8de2f6..617bd56 100644
--- a/arch/x86/include/asm/desc.h
+++ b/arch/x86/include/asm/desc.h
@@ -288,7 +288,7 @@ static inline void load_LDT(mm_context_t *pc)
 
 static inline unsigned long get_desc_base(const struct desc_struct *desc)
 {
-	return desc->base0 | ((desc->base1) << 16) | ((desc->base2) << 24);
+	return (unsigned)(desc->base0 | ((desc->base1) << 16) | ((desc->base2) << 24));
 }
 
 static inline void set_desc_base(struct desc_struct *desc, unsigned long base)
-- 
1.6.0.6


^ permalink raw reply related	[flat|nested] 3+ messages in thread

* [PATCH] x86: Make sure get_user_desc() doesn't sign extend.
@ 2009-11-04 14:11 Chris Lalancette
  0 siblings, 0 replies; 3+ messages in thread
From: Chris Lalancette @ 2009-11-04 14:11 UTC (permalink / raw)
  To: mingo
  Cc: x86, mingo, kvm, pbonzini, linux-kernel, tglx, hpa, Chris Lalancette

The current implementation of get_user_desc() sign extends
the return value because of integer promotion rules.  For
the most part, this doesn't matter, because the top bit of
base2 is usually 0.  If, however, that bit is 1, then the
entire value will be 0xffff... which is probably not what
the caller intended.  This patch casts the entire thing
to unsigned before returning, which generates almost the
same assembly as the current code but replaces the final
"cltq" (sign extend) with a "mov %eax %eax" (zero-extend).
This fixes booting certain guests under KVM.

(3rd retry, no response to the previous 2)

Signed-off-by: Chris Lalancette <clalance@redhat.com>
---
 arch/x86/include/asm/desc.h |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/arch/x86/include/asm/desc.h b/arch/x86/include/asm/desc.h
index e8de2f6..617bd56 100644
--- a/arch/x86/include/asm/desc.h
+++ b/arch/x86/include/asm/desc.h
@@ -288,7 +288,7 @@ static inline void load_LDT(mm_context_t *pc)
 
 static inline unsigned long get_desc_base(const struct desc_struct *desc)
 {
-	return desc->base0 | ((desc->base1) << 16) | ((desc->base2) << 24);
+	return (unsigned)(desc->base0 | ((desc->base1) << 16) | ((desc->base2) << 24));
 }
 
 static inline void set_desc_base(struct desc_struct *desc, unsigned long base)
-- 
1.6.0.6


^ permalink raw reply related	[flat|nested] 3+ messages in thread

* [PATCH] x86: Make sure get_user_desc() doesn't sign extend.
@ 2009-10-28 15:35 Chris Lalancette
  0 siblings, 0 replies; 3+ messages in thread
From: Chris Lalancette @ 2009-10-28 15:35 UTC (permalink / raw)
  To: mingo; +Cc: x86, mingo, kvm, pbonzini, linux-kernel, Chris Lalancette

The current implementation of get_user_desc() sign extends
the return value because of integer promotion rules.  For
the most part, this doesn't matter, because the top bit of
base2 is usually 0.  If, however, that bit is 1, then the
entire value will be 0xffff... which is probably not what
the caller intended.  This patch casts the entire thing
to unsigned before returning, which generates almost the
same assembly as the current code but replaces the final
"cltq" (sign extend) with a "mov %eax %eax" (zero-extend).
This fixes booting certain guests under KVM.

(2nd resend, since no response to the last two submissions)

Signed-off-by: Chris Lalancette <clalance@redhat.com>
---
 arch/x86/include/asm/desc.h |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/arch/x86/include/asm/desc.h b/arch/x86/include/asm/desc.h
index e8de2f6..617bd56 100644
--- a/arch/x86/include/asm/desc.h
+++ b/arch/x86/include/asm/desc.h
@@ -288,7 +288,7 @@ static inline void load_LDT(mm_context_t *pc)
 
 static inline unsigned long get_desc_base(const struct desc_struct *desc)
 {
-	return desc->base0 | ((desc->base1) << 16) | ((desc->base2) << 24);
+	return (unsigned)(desc->base0 | ((desc->base1) << 16) | ((desc->base2) << 24));
 }
 
 static inline void set_desc_base(struct desc_struct *desc, unsigned long base)
-- 
1.6.0.6


^ permalink raw reply related	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2009-11-05 10:48 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2009-11-05 10:47 [PATCH] x86: Make sure get_user_desc() doesn't sign extend Chris Lalancette
  -- strict thread matches above, loose matches on Subject: below --
2009-11-04 14:11 Chris Lalancette
2009-10-28 15:35 Chris Lalancette

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.