I asked the question below a while back but got no answer.
Does that mean BCM5754 integrated GbE will not be supported
by 6.2 release at all? The HARDWARE.TXT file does not list
it, but the source code seems to have the BCM5754 support,
or at least so the CVS comments say.
Does anyone know when if_bge will support BCM5754 GbE? We
really hope to run FreeBSD on our Dell Precision desktops.
Thanks in advance!
Sincerely,
Edward
Yi-Hua Edward Yang wrote:
> Hi,
>
> I install FreeBSD 6.1 on a new Dell Precision 390,
> which has an on-board BCM5754 GbE NIC that was not
> recognized. A quick look into the CVS shows that its
> support is added at if_bge.c rev. 1.91.2.15, later
> than what is found in 6.1.
>
> So I performed cvsup, rebuilt world and kernel with
> if_bge as a separate module. Loading the module after
> reboot put kernel to fatal trap 19. Furthermore, the
> device "recognized" by the module is wrong - BCM5787
> instead of BCM5754. (I checked with pciconf that the
> on-board GbE does have the device ID of BCM5754.)
>
> Below is the fatal trap message -
>
> bge0: <Broadcom unknown BCM5787, ASIC rev. 0xb002> mem
0xecef0000-0xecefffff irq 17 at device 0.0 on pci4
> kernel trap 19 with interrupts disabled
>
> Fatal trap 19: non-maskable interrupt trap while in kernel mode
> instruction pointer = 0x20:0xc088c873
> stack pointer = 0x28:0xe74caa48
> frame pointer = 0x28:0xe74caa54
> code segment = base 0x0, limit 0xfffff, type 0x1b
> = DPL 0, pres 1, def32 1, gran 1
> processor eflags = IOPL = 0
> current process = 547 (kldload)
> [thread pid 547 tid 100071 ]
> Stopped at 0xc088c873 = pcireg_cfgwrite+0xa3: nop
> db>
>
>
> I suspect some problem in if_bge for BCM5754 when
> merging the device driver source with HEAD after 6.1
> release, but I do not have enough knowledge to debug
> the driver. I hope someone with knowledge of if_bge
> device driver source could help. Please let me know
> if other information is needed. Thanks.
>
> Sincerely,
> Edward