dc0 acting up
Savchuk Taras
taras-s-y at mail.ru
Wed Sep 29 13:18:20 PDT 2004
On Wednesday 29 September 2004 06:09, Stephen McKay wrote:
> On Tuesday, 28th September 2004, Savchuk Taras wrote:
> >On Tuesday 28 September 2004 16:17, Jesse Marquez wrote:
> >
> >FreeBSD5.3-BETA4
> >
> >I have the same output during boot:
> >> dc0: failed to force tx and rx to idle state
> >> dc0: failed to force tx and rx to idle state
> >> dc0: failed to force tx and rx to idle state
>
> This one comes up every year or so and is my fault (more or less).
> Maybe we can fix it now that people are reporting it, even though it
> is actually pretty harmless.
>
> What hardware do you have? Ie what other dc0: lines do you find in dmesg?
>
> If you are willing to edit some code and try again, what happens when
> you comment out the entire for loop at line 1379 of sys/pci/if_dc.c, which
> looks like this:
>
> for (i = 0; i < DC_TIMEOUT; i++) {
> isr = CSR_READ_4(sc, DC_ISR);
> if (isr & DC_ISR_TX_IDLE &&
> ((isr & DC_ISR_RX_STATE) == DC_RXSTATE_STOPPED
> || (isr & DC_ISR_RX_STATE) == DC_RXSTATE_WAIT)) break;
> DELAY(10);
> }
>
> if (i == DC_TIMEOUT)
> printf("dc%d: failed to force tx and "
> "rx to idle state\n", sc->dc_unit);
>
> Does your network card still work?
I have made changes in "sys/pci/if_dc.c" as you said and have had this outout
during "buildkernel":
------------------------------------------------------------------------------------------------------------------------------------
n-long-strings -mpreferred-stack-boundary=2 -ffreestanding
-Werror /usr/src/sys/pci/if_dc.c
/usr/src/sys/pci/if_dc.c: In function `dc_setcfg':
/usr/src/sys/pci/if_dc.c:1369: warning: unused variable `i'
/usr/src/sys/pci/if_dc.c:1370: warning: unused variable `isr'
*** Error code 1
Stop in /usr/obj/usr/src/sys/T53-B6.
*** Error code 1
Stop in /usr/src.
*** Error code 1
Stop in /usr/src.
root at taras.gznet.ru#
------------------------------------------------------------------------------------------------------------------------------------
>
> The long story is that this test is probably unnecessary, though the manual
> for the original Intel 21143 requires it. If it was skipped for non-Intel
> chips, I think everyone would be happy. It would be nice to have some
> experimental verification of this theory though.
>
> Stephen.
--
Taras
More information about the freebsd-current
mailing list