Hi Edgar, On 10/09/2017 10:21 AM, Edgar E. Iglesias wrote: > On Fri, Sep 22, 2017 at 02:13:16PM -0300, Philippe Mathieu-Daudé wrote: >> Hi, >> >> I have a follow up series using multiples PHY on the MDIO bus based on this >> series. > > Hi Philippe! > > I think this is a good improvement compared to todays state. > It may make sense to have the generic mdio bus functions in mdio.c > and specific phy models in separate files, thoughts?
I'm sorry I missed your mail and noticed it today since this thread got awakened with Alistair's reviews. I'll raise this series priority in my TODO and respin with Grant correct S-o-b, Alistair comments addressed and split MDIO/PHY as you suggested. After spending 2 months with the SD bus, I now fill more confident to rework the MDIO bus and think of unit testing and tracing. Regards, Phil. >> >> Grant's previous work: >> http://lists.nongnu.org/archive/html/qemu-devel/2013-02/msg00257.html >> >> "There is more work to be done, particularly in moving to the common GPIO >> api, >> but that work can be done as a follow on patch series." >> >> Grant Likely (7): >> hw/mdio: Generalize etraxfs MDIO bitbanging emulation >> hw/mdio: Add PHY register definition >> hw/mdio: Generalize phy initialization routine >> hw/mdio: Mask out read-only bits. >> hw/mdio: Refactor bitbanging state machine >> hw/mdio: Add VMState support >> hw/mdio: Use bitbang core for smc91c111 network device >> >> include/hw/net/mdio.h | 124 +++++++++++++++++++++ >> hw/net/etraxfs_eth.c | 291 >> +----------------------------------------------- >> hw/net/mdio.c | 280 ++++++++++++++++++++++++++++++++++++++++++++++ >> hw/net/smc91c111.c | 27 ++++- >> hw/net/xilinx_axienet.c | 189 +------------------------------ >> hw/net/Makefile.objs | 2 + >> 6 files changed, 438 insertions(+), 475 deletions(-) >> create mode 100644 include/hw/net/mdio.h >> create mode 100644 hw/net/mdio.c >> >> -- >> 2.14.1 >>
