* [PATCH] x86: PAT bug fix for attribute type check after reserve_memtype
@ 2008-03-24 21:39 Venki Pallipadi
2008-03-25 14:33 ` Ingo Molnar
0 siblings, 1 reply; 2+ messages in thread
From: Venki Pallipadi @ 2008-03-24 21:39 UTC (permalink / raw)
To: Ingo Molnar, Thomas Gleixner, H. Peter Anvin
Cc: linux-kernel, suresh.b.siddha
Bug fixes for reserve_memtype() call in __ioremap and pci_mmap_page_range().
If reserve_memtype returns non-zero, then it is an error and subsequent free is
not required. Requested and returned prot value check should be done when
reserve_memtype returns success.
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
---
arch/x86/mm/ioremap.c | 11 +++++++++--
arch/x86/pci/i386.c | 7 ++++++-
2 files changed, 15 insertions(+), 3 deletions(-)
Index: linux-2.6-x86.git/arch/x86/mm/ioremap.c
===================================================================
--- linux-2.6-x86.git.orig/arch/x86/mm/ioremap.c 2008-03-21 08:28:47.000000000 -0700
+++ linux-2.6-x86.git/arch/x86/mm/ioremap.c 2008-03-24 06:59:09.000000000 -0700
@@ -126,6 +126,7 @@ static void __iomem *__ioremap(unsigned
unsigned long new_prot_val;
pgprot_t prot;
void __iomem *ret_addr;
+ int retval;
/* Don't allow wraparound or zero size */
last_addr = phys_addr + size - 1;
@@ -165,8 +166,14 @@ static void __iomem *__ioremap(unsigned
phys_addr &= PAGE_MASK;
size = PAGE_ALIGN(last_addr+1) - phys_addr;
- if (reserve_memtype(phys_addr, phys_addr + size,
- prot_val, &new_prot_val)) {
+ retval = reserve_memtype(phys_addr, phys_addr + size,
+ prot_val, &new_prot_val);
+ if (retval) {
+ printk("reserve_memtype returned %d\n", retval);
+ return NULL;
+ }
+
+ if (prot_val != new_prot_val) {
/*
* Do not fallback to certain memory types with certain
* requested type:
Index: linux-2.6-x86.git/arch/x86/pci/i386.c
===================================================================
--- linux-2.6-x86.git.orig/arch/x86/pci/i386.c 2008-03-21 08:28:47.000000000 -0700
+++ linux-2.6-x86.git/arch/x86/pci/i386.c 2008-03-24 07:01:22.000000000 -0700
@@ -328,6 +328,7 @@ int pci_mmap_page_range(struct pci_dev *
unsigned long len = vma->vm_end - vma->vm_start;
unsigned long flags;
unsigned long new_flags;
+ int retval;
/* I/O space cannot be accessed via normal processor loads and
* stores on this platform.
@@ -344,7 +345,11 @@ int pci_mmap_page_range(struct pci_dev *
vma->vm_page_prot = __pgprot(prot);
flags = pgprot_val(vma->vm_page_prot) & _PAGE_CACHE_MASK;
- if (reserve_memtype(addr, addr + len, flags, &new_flags)) {
+ retval = reserve_memtype(addr, addr + len, flags, &new_flags);
+ if (retval)
+ return retval;
+
+ if (flags != new_flags) {
/*
* Do not fallback to certain memory types with certain
* requested type:
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [PATCH] x86: PAT bug fix for attribute type check after reserve_memtype
2008-03-24 21:39 [PATCH] x86: PAT bug fix for attribute type check after reserve_memtype Venki Pallipadi
@ 2008-03-25 14:33 ` Ingo Molnar
0 siblings, 0 replies; 2+ messages in thread
From: Ingo Molnar @ 2008-03-25 14:33 UTC (permalink / raw)
To: Venki Pallipadi
Cc: Thomas Gleixner, H. Peter Anvin, linux-kernel, suresh.b.siddha
* Venki Pallipadi <venkatesh.pallipadi@intel.com> wrote:
> Bug fixes for reserve_memtype() call in __ioremap and
> pci_mmap_page_range(). If reserve_memtype returns non-zero, then it is
> an error and subsequent free is not required. Requested and returned
> prot value check should be done when reserve_memtype returns success.
thanks, applied.
> + if (retval) {
> + printk("reserve_memtype returned %d\n", retval);
that needs a KERN_INFO, and should be labeled "Warning: ...".
Ingo
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2008-03-25 14:33 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-03-24 21:39 [PATCH] x86: PAT bug fix for attribute type check after reserve_memtype Venki Pallipadi
2008-03-25 14:33 ` Ingo Molnar
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox