On Mon, Jun 12, 2023 at 3:01 PM Igor Mammedov <imamm...@redhat.com> wrote:
> On Sun, 11 Jun 2023 12:33:59 +0200 > Bernhard Beschow <shen...@gmail.com> wrote: > > > Fixes the following clangd warning (-Winitializer-overrides): > > > > q35.c:297:19: Initializer overrides prior initialization of this > subobject > > q35.c:292:19: previous initialization is here > > > > Settle on native endian which causes the least overhead. > indeed it doesn't matter which way we read all ones, so that should work. > but does it really matter (I mean the overhead/what workload)? > If not, I'd prefer explicit LE as it's now to be consistent > the the rest of memops on Q35. > I got a comment from Michael about this in [1], so I've changed it. I don't mind changing it either way. Best regards, Bernhard [1] https://patchew.org/QEMU/20230214131441.101760-1-shen...@gmail.com/20230214131441.101760-3-shen...@gmail.com/#20230301164339-mutt-send-email-...@kernel.org > > > > > Fixes: bafc90bdc594 ("q35: implement TSEG") > > Signed-off-by: Bernhard Beschow <shen...@gmail.com> > > --- > > hw/pci-host/q35.c | 1 - > > 1 file changed, 1 deletion(-) > > > > diff --git a/hw/pci-host/q35.c b/hw/pci-host/q35.c > > index fd18920e7f..859c197f25 100644 > > --- a/hw/pci-host/q35.c > > +++ b/hw/pci-host/q35.c > > @@ -290,7 +290,6 @@ static const MemoryRegionOps blackhole_ops = { > > .valid.max_access_size = 4, > > .impl.min_access_size = 4, > > .impl.max_access_size = 4, > > - .endianness = DEVICE_LITTLE_ENDIAN, > > }; > > > > /* PCIe MMCFG */ > >
