The subject "pci: use uint64_t for unsigned long values" does not reflect what the patch does. This should be tagged v2. Generally speaking, resend the whole series as v2 (or v3, etc...) even if only one patch is being tweaked. It's fairly obvious in this case that the other two other patches from v1 are still relevant, but that won't always be true. On Sat, Oct 12, 2019 at 01:26:23AM -0700, Bill Wendling wrote: > The "pci_bar_*" functions use 64-bit masks, but the results are assigned > to 32-bit variables. Use 32-bit masks, since we're interested only in > the least significant 4-bits. > > Signed-off-by: Bill Wendling <morbo@xxxxxxxxxx> > --- > lib/linux/pci_regs.h | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/lib/linux/pci_regs.h b/lib/linux/pci_regs.h > index 1becea8..3bc2b92 100644 > --- a/lib/linux/pci_regs.h > +++ b/lib/linux/pci_regs.h > @@ -96,8 +96,8 @@ > #define PCI_BASE_ADDRESS_MEM_TYPE_1M 0x02 /* Below 1M [obsolete] */ > #define PCI_BASE_ADDRESS_MEM_TYPE_64 0x04 /* 64 bit address */ > #define PCI_BASE_ADDRESS_MEM_PREFETCH 0x08 /* prefetchable? */ > -#define PCI_BASE_ADDRESS_MEM_MASK (~0x0fUL) > -#define PCI_BASE_ADDRESS_IO_MASK (~0x03UL) > +#define PCI_BASE_ADDRESS_MEM_MASK (~0x0fU) > +#define PCI_BASE_ADDRESS_IO_MASK (~0x03U) The defines themselves should still be 64-bit, e.g. there's a flag right above to denote a 64-bit address. Explicitly casting the return value in pci_bar_mask() where we're dealing with 32-bit addresses would be preferable. > /* bit 1 is reserved if address_space = 1 */ > > /* Header type 0 (normal devices) */ > -- > 2.23.0.700.g56cf767bdb-goog >
