| Message ID | 1669138842-30100-1-git-send-email-mikelley@microsoft.com (mailing list archive) |
|---|---|
| State | Not Applicable |
| Headers | show |
| Series | [v4,1/1] x86/ioremap: Fix page aligned size calculation in __ioremap_caller() | expand |
| Context | Check | Description |
|---|---|---|
| netdev/tree_selection | success | Not a local patch |
On Tue, Nov 22, 2022 at 09:40:42AM -0800, Michael Kelley wrote: > Current code re-calculates the size after aligning the starting and > ending physical addresses on a page boundary. But the re-calculation > also embeds the masking of high order bits that exceed the size of > the physical address space (via PHYSICAL_PAGE_MASK). If the masking > removes any high order bits, the size calculation results in a huge > value that is likely to immediately fail. > > Fix this by re-calculating the page-aligned size first. Then mask any > high order bits using PHYSICAL_PAGE_MASK. > > Fixes: ffa71f33a820 ("x86, ioremap: Fix incorrect physical address handling in PAE mode") > Acked-by: Dave Hansen <dave.hansen@linux.intel.com> > Signed-off-by: Michael Kelley <mikelley@microsoft.com> Reviewed-by: Wei Liu <wei.liu@kernel.org> > --- > > This patch was previously Patch 1 of a larger series[1]. Breaking > it out separately per discussion with Dave Hansen and Boris Petkov. > > [1] https://lore.kernel.org/linux-hyperv/1668624097-14884-1-git-send-email-mikelley@microsoft.com/ > > arch/x86/mm/ioremap.c | 8 +++++++- > 1 file changed, 7 insertions(+), 1 deletion(-) > > diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c > index 78c5bc6..6453fba 100644 > --- a/arch/x86/mm/ioremap.c > +++ b/arch/x86/mm/ioremap.c > @@ -217,9 +217,15 @@ static void __ioremap_check_mem(resource_size_t addr, unsigned long size, > * Mappings have to be page-aligned > */ > offset = phys_addr & ~PAGE_MASK; > - phys_addr &= PHYSICAL_PAGE_MASK; > + phys_addr &= PAGE_MASK; > size = PAGE_ALIGN(last_addr+1) - phys_addr; > > + /* > + * Mask out any bits not part of the actual physical > + * address, like memory encryption bits. > + */ > + phys_addr &= PHYSICAL_PAGE_MASK; > + > retval = memtype_reserve(phys_addr, (u64)phys_addr + size, > pcm, &new_pcm); > if (retval) { > -- > 1.8.3.1 >
From: Wei Liu <wei.liu@kernel.org> Sent: Friday, November 25, 2022 7:20 AM > > On Tue, Nov 22, 2022 at 09:40:42AM -0800, Michael Kelley wrote: > > Current code re-calculates the size after aligning the starting and > > ending physical addresses on a page boundary. But the re-calculation > > also embeds the masking of high order bits that exceed the size of > > the physical address space (via PHYSICAL_PAGE_MASK). If the masking > > removes any high order bits, the size calculation results in a huge > > value that is likely to immediately fail. > > > > Fix this by re-calculating the page-aligned size first. Then mask any > > high order bits using PHYSICAL_PAGE_MASK. > > > > Fixes: ffa71f33a820 ("x86, ioremap: Fix incorrect physical address handling in PAE mode") > > Acked-by: Dave Hansen <dave.hansen@linux.intel.com> > > Signed-off-by: Michael Kelley <mikelley@microsoft.com> > > Reviewed-by: Wei Liu <wei.liu@kernel.org> > > > --- > > > > This patch was previously Patch 1 of a larger series[1]. Breaking > > it out separately per discussion with Dave Hansen and Boris Petkov. > > > > [1] https://lore.kernel.org/linux-hyperv/1668624097-14884-1-git-send-email-mikelley@microsoft.com/ > > Boris -- you were going to pick up this patch separately though urgent. Can you go ahead and do that? https://lore.kernel.org/linux-hyperv/Y3vo5drAFPQSsrF4@zn.tnic/ Michael > > arch/x86/mm/ioremap.c | 8 +++++++- > > 1 file changed, 7 insertions(+), 1 deletion(-) > > > > diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c > > index 78c5bc6..6453fba 100644 > > --- a/arch/x86/mm/ioremap.c > > +++ b/arch/x86/mm/ioremap.c > > @@ -217,9 +217,15 @@ static void __ioremap_check_mem(resource_size_t addr, unsigned long size, > > * Mappings have to be page-aligned > > */ > > offset = phys_addr & ~PAGE_MASK; > > - phys_addr &= PHYSICAL_PAGE_MASK; > > + phys_addr &= PAGE_MASK; > > size = PAGE_ALIGN(last_addr+1) - phys_addr; > > > > + /* > > + * Mask out any bits not part of the actual physical > > + * address, like memory encryption bits. > > + */ > > + phys_addr &= PHYSICAL_PAGE_MASK; > > + > > retval = memtype_reserve(phys_addr, (u64)phys_addr + size, > > pcm, &new_pcm); > > if (retval) { > > -- > > 1.8.3.1 > >
On Mon, Nov 28, 2022 at 02:43:28PM +0000, Michael Kelley (LINUX) wrote: > Boris -- you were going to pick up this patch separately > though urgent. Can you go ahead and do that? Did you not get the tip-bot notification? https://lore.kernel.org/r/166911713030.4906.16935727667401525991.tip-bot2@tip-bot2 you're on Cc there too.
From: Borislav Petkov <bp@alien8.de> Sent: Monday, November 28, 2022 7:46 AM > > On Mon, Nov 28, 2022 at 02:43:28PM +0000, Michael Kelley (LINUX) wrote: > > Boris -- you were going to pick up this patch separately > > though urgent. Can you go ahead and do that? > > Did you not get the tip-bot notification? > > https://lore.kernel.org/all/166911713030.4906.16935727667401525991.tip-bot2@tip-bot2/ > > you're on Cc there too. > Argh. My mistake. Thanks. Michael
diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c index 78c5bc6..6453fba 100644 --- a/arch/x86/mm/ioremap.c +++ b/arch/x86/mm/ioremap.c @@ -217,9 +217,15 @@ static void __ioremap_check_mem(resource_size_t addr, unsigned long size, * Mappings have to be page-aligned */ offset = phys_addr & ~PAGE_MASK; - phys_addr &= PHYSICAL_PAGE_MASK; + phys_addr &= PAGE_MASK; size = PAGE_ALIGN(last_addr+1) - phys_addr; + /* + * Mask out any bits not part of the actual physical + * address, like memory encryption bits. + */ + phys_addr &= PHYSICAL_PAGE_MASK; + retval = memtype_reserve(phys_addr, (u64)phys_addr + size, pcm, &new_pcm); if (retval) {
Current code re-calculates the size after aligning the starting and ending physical addresses on a page boundary. But the re-calculation also embeds the masking of high order bits that exceed the size of the physical address space (via PHYSICAL_PAGE_MASK). If the masking removes any high order bits, the size calculation results in a huge value that is likely to immediately fail. Fix this by re-calculating the page-aligned size first. Then mask any high order bits using PHYSICAL_PAGE_MASK. Fixes: ffa71f33a820 ("x86, ioremap: Fix incorrect physical address handling in PAE mode") Acked-by: Dave Hansen <dave.hansen@linux.intel.com> Signed-off-by: Michael Kelley <mikelley@microsoft.com> --- This patch was previously Patch 1 of a larger series[1]. Breaking it out separately per discussion with Dave Hansen and Boris Petkov. [1] https://lore.kernel.org/linux-hyperv/1668624097-14884-1-git-send-email-mikelley@microsoft.com/ arch/x86/mm/ioremap.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)