Hi Stefano, > On 3 Sep 2022, at 12:21 am, Stefano Stabellini <sstabell...@kernel.org> wrote: > > On Fri, 2 Sep 2022, Rahul Singh wrote: >> From: Zhou Wang <wangzh...@hisilicon.com> >> >> Backport Linux commit a76a37777f2c. This is the clean backport without >> any changes. >> >> Reading the 'prod' MMIO register in order to determine whether or >> not there is valid data beyond 'cons' for a given queue does not >> provide sufficient dependency ordering, as the resulting access is >> address dependent only on 'cons' and can therefore be speculated >> ahead of time, potentially allowing stale data to be read by the >> CPU. >> >> Use readl() instead of readl_relaxed() when updating the shadow copy >> of the 'prod' pointer, so that all speculated memory reads from the >> corresponding queue can occur only from valid slots. >> >> Signed-off-by: Zhou Wang <wangzh...@hisilicon.com> >> Link: >> https://lore.kernel.org/r/1601281922-117296-1-git-send-email-wangzh...@hisilicon.com >> [will: Use readl() instead of explicit barrier. Update 'cons' side to match.] >> Signed-off-by: Will Deacon <w...@kernel.org> >> Origin: git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git >> a76a37777f2c >> Signed-off-by: Rahul Singh <rahul.si...@arm.com> >> --- >> Changes in v2: >> - fix commit msg >> - add _iomb changes also from the origin patch >> --- >> xen/arch/arm/include/asm/system.h | 1 + >> xen/drivers/passthrough/arm/smmu-v3.c | 11 +++++++++-- >> 2 files changed, 10 insertions(+), 2 deletions(-) >> >> diff --git a/xen/arch/arm/include/asm/system.h >> b/xen/arch/arm/include/asm/system.h >> index 65d5c8e423..fe27cf8c5e 100644 >> --- a/xen/arch/arm/include/asm/system.h >> +++ b/xen/arch/arm/include/asm/system.h >> @@ -29,6 +29,7 @@ >> #endif >> >> #define smp_wmb() dmb(ishst) >> +#define __iomb() dmb(osh) > > We don't have any other #define starting with __ in system.h. > I wonder if we should call this macro differently or simply iomb(). I think either iomb() or dma_mb() will be the right name. Please let me know your view on this.
Regards, Rahul
