tech-net archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Another update for axe(4)



hi,

I'm sorry my work is incomplete.

I checked your patch.

- it should not ignore reported phy address. I have the device which can
read a value from a register via *invalid* phy address, it results
invalid phy address is used to access phy.

- it should add a quirk for devices which lies about phy address. but I
don't have such a device so I can't test it...

- in axe_setmulti(), "if (ifp->if_flags & IFF_ALLMULTI) {" must be "...
IFF_PROMISC ...".

- in axe_attach(), no need to read AXE_CMD_READ_PHYID twice.

- in axe_encap(), USBD_NO_COPY can be specified in usbd_setup_xfer().

- in axe_init(), splnet() should be called before using ifp->if_flags.

- in axe_init(), no need to set AXE_RXCMD_PROMISC and
AXE_RXCMD_BROADCAST individually.

- in axe_init(), usbd_setup_xfer(AXE_ENDPT_RX) should use c->axe_buf
instead of mtod(c->axe_mbuf). c->axe_mbuf seems a placeholder of unused
mbuf. received data should go into c->axe_buf. see axe_rxeof().

- in axe_ioctl(), call ifioctl_common() at first in SIOCSIFFLAGS case.

- in axe_watchdog(), "if (IFQ_IS_EMPTY(&ifp->if_snd))" must be
!IFQ_IS_EMPTY().

Regards,

--
FUKAUMI Naoki


Home | Main Index | Thread Index | Old Index