On 26/10/17 11:12, Mark Cave-Ayland wrote: > On 25/10/17 18:47, Peter Maydell wrote: > >> On 25 October 2017 at 16:59, Mark Cave-Ayland >> <mark.cave-ayl...@ilande.co.uk> wrote: >>> This enables them to be used outside of lance.c. >>> >>> Signed-off-by: Mark Cave-Ayland <mark.cave-ayl...@ilande.co.uk> >>> CC: Jason Wang <jasow...@redhat.com> >> >> >>> diff --git a/include/hw/net/lance.h b/include/hw/net/lance.h >>> new file mode 100644 >>> index 0000000..c112fbc >>> --- /dev/null >>> +++ b/include/hw/net/lance.h >>> @@ -0,0 +1,41 @@ >>> +/* >>> + * QEMU AMD PC-Net II (Am79C970A) emulation >> >> Filename says this is Lance ethernet, but the comment says it's PC-Net ? > > According to the datasheet for Am79C970A, the original Lance is an > Am7990 device and the Am79C970A aka PCNet-PCI II as emulated by QEMU is > register-compatible with it. > > I guess the comment above is more technically correct, but I'm happy to > adjust it in my local tree if you still feel it needs to change? > >>> +#define TYPE_LANCE "lance" >>> +#define SYSBUS_PCNET(obj) \ >>> + OBJECT_CHECK(SysBusPCNetState, (obj), TYPE_LANCE) >> >> It's a bit weird that the OBJECT_CHECK macro name >> and the state structure name don't line up with >> the device name, but this is just code motion so >> not a problem. > > Yeah, same issue as above... :/ > >> Other than fixing the comment, >> Reviewed-by: Peter Maydell <peter.mayd...@linaro.org> >> >> (In an ideal world pcnet.h would be in include/.) >> >> thanks >> -- PMM
Hi Peter, Did you have any further comments on the above? If not I can tidy up my branch for this (and the sun4m IOMMU work) later today and submit a PR in time for freeze. ATB, Mark.