Message ID | 1395396367-19110-1-git-send-email-binw@marvell.com |
---|---|
State | New |
Headers | show |
Hi Bin, On 21/03/14 10:06, Bin Wang wrote: > kernel panic happened when iommu_unmap a buffer larger than 2MB, > more than expected pmd entries got “invalidated”, due to a wrong range > passed to arm_smmu_alloc_init_pte. it was likely a typo, now we fix > it, passing the correct "end" address to arm_smmu_alloc_init_pte. > > Signed-off-by: Bin Wang <binw@marvell.com> > --- > drivers/iommu/arm-smmu.c | 2 +- > 1 files changed, 1 insertions(+), 1 deletions(-) > > diff --git a/drivers/iommu/arm-smmu.c b/drivers/iommu/arm-smmu.c > index 1d9ab39..52d0c34 100644 > --- a/drivers/iommu/arm-smmu.c > +++ b/drivers/iommu/arm-smmu.c > @@ -1346,7 +1346,7 @@ static int arm_smmu_alloc_init_pmd(struct arm_smmu_device *smmu, pud_t *pud, > > do { > next = pmd_addr_end(addr, end); > - ret = arm_smmu_alloc_init_pte(smmu, pmd, addr, end, pfn, > + ret = arm_smmu_alloc_init_pte(smmu, pmd, addr, next, pfn, > flags, stage); > phys += next - addr; > } while (pmd++, addr = next, addr < end); > Looks like a valid fix. In Will's absence ([1]), I'll queue it for 3.15, and CC stable. Joerg: are you still taking pull requests for 3.15, or shall I wait until rc1? Thanks, M. [1]: http://www.spinics.net/lists/arm-kernel/msg315615.html
On Fri, Mar 21, 2014 at 10:57:08AM +0000, Marc Zyngier wrote: > Joerg: are you still taking pull requests for 3.15, or shall I wait > until rc1? No need to wait for rc1, send me a pull-request and I will send the fix upstream. Joerg
diff --git a/drivers/iommu/arm-smmu.c b/drivers/iommu/arm-smmu.c index 1d9ab39..52d0c34 100644 --- a/drivers/iommu/arm-smmu.c +++ b/drivers/iommu/arm-smmu.c @@ -1346,7 +1346,7 @@ static int arm_smmu_alloc_init_pmd(struct arm_smmu_device *smmu, pud_t *pud, do { next = pmd_addr_end(addr, end); - ret = arm_smmu_alloc_init_pte(smmu, pmd, addr, end, pfn, + ret = arm_smmu_alloc_init_pte(smmu, pmd, addr, next, pfn, flags, stage); phys += next - addr; } while (pmd++, addr = next, addr < end);
kernel panic happened when iommu_unmap a buffer larger than 2MB, more than expected pmd entries got “invalidated”, due to a wrong range passed to arm_smmu_alloc_init_pte. it was likely a typo, now we fix it, passing the correct "end" address to arm_smmu_alloc_init_pte. Signed-off-by: Bin Wang <binw@marvell.com> --- drivers/iommu/arm-smmu.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)