[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] [RFC PATCH v1 2/2] e1000: adjust initial autoneg timing
From: |
Michael S. Tsirkin |
Subject: |
Re: [Qemu-devel] [RFC PATCH v1 2/2] e1000: adjust initial autoneg timing (for piix/osx) |
Date: |
Wed, 2 Jul 2014 18:17:05 +0300 |
On Wed, Jul 02, 2014 at 10:21:34AM -0400, Gabriel L. Somlo wrote:
> On Wed, Jul 02, 2014 at 12:33:16PM +0300, Michael S. Tsirkin wrote:
> > Just poking around the spec I find more things
> > we don't implement correctly wrt to auto-negotiation.
> > For example, MII_SR_AUTONEG_CAPS isn't set, is it?
> > Maybe that's why your guest doesn't work:
> > it doesn't expect to get autonegotation at all?
> >
> > So I have a question: does your patch actually help any guests?
> > If not, maybe we should defer it to after release,
> > and try to clean up autonegotiation more thouroughly for 2.2?
>
> I'll re-submit after 2.1 is officially out. But, since we're talking
> about MII_SR_AUTONEG_CAPS: PHY_STATUS is initialized to 0x794d, which
> includes setting the MII_SR_AUTONEG_CAPS bit (|= 0x8).
> Did you mean: we should check for MII_SR_AUTONEG_CAPS in have_autoneg() ?
> (i.e., on the chance it gets turned off by a guest-side write to
> PHY_STATUS) ?
PHY_STATUS isn't writeable is it?
No I just got confused with the binary math.
We really should use symbolic constants there,
we already have them defined.
>
> Thx,
> --G
>
> PS. Maybe also spell out the individual bits in phy_reg_init[] ? Like,
> instead of:
>
> [PHY_STATUS] = 0x794d,
>
> do this:
>
> [PHY_STATUS] = MII_SR_EXTENDED_CAPS |
> MII_SR_LINK_STATUS |
> MII_SR_AUTONEG_CAPS |
> MII_SR_PREAMBLE_SUPPRESS |
> MII_SR_EXTENDED_STATUS |
> MII_SR_10T_HD_CAPS |
> MII_SR_10T_FD_CAPS |
> MII_SR_100X_HD_CAPS |
> MII_SR_100X_FD_CAPS,
>
> ... for all registers ? Much more verbose, but IMHO that'd be a good
> thing :)
>
>
- Re: [Qemu-devel] e1000 autoneg timing, piix/osx, (continued)
- Re: [Qemu-devel] e1000 autoneg timing, piix/osx, Gabriel L. Somlo, 2014/07/03
- Re: [Qemu-devel] e1000 autoneg timing, piix/osx, Alexander Graf, 2014/07/03
- Re: [Qemu-devel] e1000 autoneg timing, piix/osx, Alexander Graf, 2014/07/03
- Re: [Qemu-devel] e1000 autoneg timing, piix/osx, Paolo Bonzini, 2014/07/03
- Re: [Qemu-devel] e1000 autoneg timing, piix/osx, Gabriel L. Somlo, 2014/07/03
- Re: [Qemu-devel] e1000 autoneg timing, piix/osx, Alexander Graf, 2014/07/03
Re: [Qemu-devel] [RFC PATCH v1 2/2] e1000: adjust initial autoneg timing (for piix/osx), Michael S. Tsirkin, 2014/07/02