Urgent memory fix for kvm with unaligned memory slots
The memory core may generate RAM memory regions that are not page aligned, but the kvm code is not prepared to handle them well and will abort under certain conditions. This patch fixes the problem.
Please pull from:
git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git memory/urgent
----------------------------------------------------------------
Avi Kivity (1):
kvm: fix unaligned slots
kvm-all.c | 15 ++++++++++++--- 1 files changed, 12 insertions(+), 3 deletions(-)
kvm-all.c +0 -0
@@ -541,17 +541,26 @@ static void kvm_set_phys_mem(MemoryRegionSection
*section, bool add)
target_phys_addr_t start_addr = section->offset_within_address_space;
ram_addr_t size = section->size;
void *ram = NULL;+ unsigned delta;
/* kvm works in page size chunks, but the function may be called
with sub-page size and unaligned start address. */
- size = TARGET_PAGE_ALIGN(size);
- start_addr = TARGET_PAGE_ALIGN(start_addr);
+ delta = TARGET_PAGE_ALIGN(size) - size;
+ if (delta > size) {
+ return;
+ }
+ start_addr += delta;
+ size -= delta;
+ size &= TARGET_PAGE_MASK;
+ if (!size || (start_addr & ~TARGET_PAGE_MASK)) {+ return;
+ }
if (!memory_region_is_ram(mr)) {
return;
}- ram = memory_region_get_ram_ptr(mr) + section->offset_within_region;
+ ram = memory_region_get_ram_ptr(mr) + section->offset_within_region
+ delta;
while (1) {
mem = kvm_lookup_overlapping_slot(s, start_addr, start_addr +
size);
Signature · 1 line
error compiling committee.c: too many arguments to function
Quoted 41 lines · Avi Kivity
The memory core may generate RAM memory regions that are not page aligned, but the kvm code is not prepared to handle them well and will abort under certain conditions. This patch fixes the problem.
Please pull from:
git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git memory/urgent
----------------------------------------------------------------
Avi Kivity (1):
kvm: fix unaligned slotskvm-all.c | 15 ++++++++++++---
1 files changed, 12 insertions(+), 3 deletions(-)kvm-all.c +1
-0@@ -541,17 +541,26 @@ static void kvm_set_phys_mem(MemoryRegionSection *section, bool add) target_phys_addr_t start_addr = section->offset_within_address_space; ram_addr_t size = section->size; void *ram = NULL; + unsigned delta; /* kvm works in page size chunks, but the function may be calledwith sub-page size and unaligned start address. */
- size =TARGET_PAGE_ALIGN(size);
- start_addr =TARGET_PAGE_ALIGN(start_addr);
+ delta =TARGET_PAGE_ALIGN(size)- size;
+ if (delta > size) {
+ return;
+ }
+ start_addr += delta;
+ size -= delta;
+ size &=TARGET_PAGE_MASK;
+ if (!size || (start_addr& ~TARGET_PAGE_MASK)) {
+ return;
+ }
Quoted 7 lines
if (!memory_region_is_ram(mr)) {
return;
}- ram =
memory_region_get_ram_ptr(mr)+section->offset_within_region;
+ ram =memory_region_get_ram_ptr(mr)+section->offset_within_region
+ delta;
Am I crazy, or does this look wrong?
Quoted 12 lines
while (1) {
mem =kvm_lookup_overlapping_slot(s, start_addr, start_addr + size);--
error compilingcommittee.c: too many arguments to function--
To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Quoted 19 lines · Bobby Powers
Quoted 19 lines
/* kvm works in page size chunks, but the function may be called
with sub-page size and unaligned start address. */(quoted diff) +10
-2- size = TARGET_PAGE_ALIGN(size); - start_addr = TARGET_PAGE_ALIGN(start_addr); + delta = TARGET_PAGE_ALIGN(size) - size; + if (delta > size) { + return; + } + start_addr += delta; + size -= delta; + size &= TARGET_PAGE_MASK; + if (!size || (start_addr & ~TARGET_PAGE_MASK)) { + return; + } if (!memory_region_is_ram(mr)) { return; }
Quoted 5 lines
Review
Quoted 3 lines
(quoted diff) +2
-1- ram = memory_region_get_ram_ptr(mr) + section->offset_within_region; + ram = memory_region_get_ram_ptr(mr) + section->offset_within_region + delta;Am I crazy, or does this look wrong?
Could be both. Why do you thing it is wrong?
Signature · 1 line
error compiling committee.c: too many arguments to function
Quoted 8 lines · Avi Kivity
Quoted 6 lines
Review
Quoted 4 lines
(quoted diff) +2
-1- ram = memory_region_get_ram_ptr(mr) + section->offset_within_region; + ram = memory_region_get_ram_ptr(mr) + section->offset_within_region + delta;Am I crazy, or does this look wrong?
Could be both. Why do you thing it is wrong?
Line wrapping makes it look like we are adding two lines, one line ending in 'section->offset_within_region', and the next line starting with 'delta;', which is a syntax error.
But without line wrapping, we are adding just one line with 'offset_within_region + delta;' at the end of that line.
Signature · 2 lines
Eric Blake [email protected] +1-919-301-3266 Libvirt virtualization library http://libvirt.org
Quoted 6 lines · Eric Blake
Quoted 5 lines · Avi Kivity
Quoted 5 lines
Quoted 4 lines
- ram =
memory_region_get_ram_ptr(mr)+section->offset_within_region;
+ ram =memory_region_get_ram_ptr(mr)+section->offset_within_region
+ delta;
Quoted 7 lines
Quoted 3 lines
Quoted 1 line
Am I crazy, or does this look wrong?
Could be both. Why do you thing it is wrong?
Line wrapping makes it look like we are adding two lines, one line ending in '
section->offset_within_region', and the next line starting with 'delta;', which is a syntax error.
yea, patch line wrapping was making my eyes see things. okay, sorry for the noise.
Quoted 8 lines
But without line wrapping, we are adding just one line with '
offset_within_region+ delta;' at the end of that line.--
Eric Blake [email protected] +1-919-301-3266
Libvirt virtualization library http://libvirt.org
Quoted 8 lines · Eric Blake
Quoted 7 lines · Avi Kivity
Quoted 6 lines
Review
Quoted 4 lines
(quoted diff) +2
-1- ram = memory_region_get_ram_ptr(mr) + section->offset_within_region; + ram = memory_region_get_ram_ptr(mr) + section->offset_within_region + delta;Am I crazy, or does this look wrong?
Quoted 9 lines
Quoted 1 line
Could be both. Why do you thing it is wrong?
Line wrapping makes it look like we are adding two lines, one line ending in '
section->offset_within_region', and the next line starting with 'delta;', which is a syntax error.But without line wrapping, we are adding just one line with '
offset_within_region+ delta;' at the end of that line.
Ah, of course. I just copy/pasted this into thunderbird, this is not a proper patch but a pull request. Sorry about the confusion.
Bobby: so it does indeed look wrong. This says nothing about your sanity though, for that consult a qualified professional instead of asking on the mailing list.
Signature · 1 line
error compiling committee.c: too many arguments to function
Quoted 7 lines · Avi Kivity
The memory core may generate RAM memory regions that are not page aligned, but the kvm code is not prepared to handle them well and will abort under certain conditions. This patch fixes the problem.
Please pull from:
git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.gitmemory/urgent
Pulled. Thanks.
Regards,
Anthony Liguori
Quoted 46 lines
----------------------------------------------------------------
Avi Kivity (1):
kvm: fix unaligned slotskvm-all.c | 15 ++++++++++++--- 1 files changed, 12 insertions(+), 3 deletions(-)kvm-all.c +13
-3@@ -541,17 +541,26 @@ static void kvm_set_phys_mem(MemoryRegionSection *section, bool add) target_phys_addr_t start_addr = section->offset_within_address_space;ram_addr_t size = section->size;void *ram =NULL ; +unsigned delta;/* kvm works in page size chunks, but the function may be called with sub-page size and unaligned start address. */ - size =TARGET_PAGE_ALIGN (size); - start_addr =TARGET_PAGE_ALIGN (start_addr); + delta =TARGET_PAGE_ALIGN (size) - size; +if (delta> size) { +return ; + } + start_addr += delta; + size -= delta; + size&=TARGET_PAGE_MASK ; +if (!size || (start_addr& ~TARGET_PAGE_MASK )) { +return ; + }if (!memory_region_is_ram (mr)) {return ; } - ram =memory_region_get_ram_ptr (mr) + section->offset_within_region; + ram =memory_region_get_ram_ptr (mr) + section->offset_within_region + delta;while (1 ) { mem = kvm_lookup_overlapping_slot(s, start_addr, start_addr +size);