Проблемма с LAN (PHY MII/RMII) : at91sam9g20ek+PHY Davicom DM9161AEP, ifconfig : error overruns Опции

[salara]

Добрый вечер. Есть проблемма - помогите, пожалуйста, решить.
Суть ее такова :
Есть плата at91sam9g20ek с PHY Davicom DM9161AEP.

с ядром linux 2.6.30 вот такая беда твориться :

kernel BUG at drivers/net/macb.c:442!
Unable to handle kernel NULL pointer dereference at virtual address 00000000
pgd = c79e4000
[00000000] *pgd=279fa031, *pte=00000000, *ppte=00000000
Internal error: Oops: 817 [#1]
Modules linked in: vin_mod dss_mod k16_mod scsi_wait_scan
CPU: 0 Not tainted (2.6.30 #40)
PC is at __bug+0x1c/0x28
LR is at __bug+0x18/0x28
pc : [<c002974c>] lr : [<c0029748>] psr: 60000013
sp : c79e1e58 ip : fefff200 fp : 00000040
r10: 00000100 r9 : c78ebb38 r8 : 00000121
r7 : 000000d6 r6 : c78ebb00 r5 : c782e4e0 r4 : 00000120
r3 : 00000000 r2 : 80000013 r1 : 000019c1 r0 : 00000029
Flags: nZCv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user
Control: 0005317f Table: 279e4000 DAC: 00000015
Process elgatos (pid: 497, stack limit = 0xc79e0268)
Stack: (0xc79e1e58 to 0xc79e2000)
1e40: 0000011f c0185ee8
1e60: 00000040 00000000 c79e0000 c78ebb38 00000040 00000100 0000012c c031410c
......
......

перед этим ifconfig мне показывал ошибки overruns :

eth0 Link encap:Ethernet HWaddr 00:12:34:56:09:8F
inet addr:192.100.101.254 Bcast:192.100.101.255 Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:170240 errors:82 dropped:0 overruns:82 frame:0
TX packets:188302 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:31242471 (29.7 MiB) TX bytes:40814141 (38.9 MiB)
Interrupt:21 Base address:0x4000

когда грузилось ядро было такое сообщение :

MACB_mii_bus: probed
eth0: Atmel MACB at 0xfffc4000 irq 21 (00:12:34:56:10:8f)
eth0: attached PHY driver [Davicom DM9161A] (mii_bus:phy_addr=ffffffff:00, irq=-1)

плюнул на это ядро и перешел на 2.6.33 (с патчем от linux4sam),
ядро больше не падает, но ошибки overruns остались, причем их количество
растет при увеличении трафика (sip, rtp-пакеты) через LAN.
Подключил плату к свитчу с портом в 10Mbps/Full-Duplex - ошибки overruns исчезли.
Однако на плате не могу принудительно установить режим 10/Full-Duplex:

mii-tool говорит :
SIOCGMIIPHY on 'eth0' failed: Operation not supported
no MII interfaces found
примерно тоже говорит ifconfig при попытке сменить media у LAN-а

Вопрос такой :
режим MII не поддерживается драйвером ядра macb для PHY Davicom DM9161AEP ????????????
может кто-то с такой проблеммой уже сталкивался (что-то подобное читал у miniMax-а, но почему-то
его рекомеднации мне не помогли).

Кто вкурсе, помогите, пожалуйста.

[salara]

Уважаемые, помогите кто в курсе этой беды.
теперь и на ядре 2.6.33 тот же баг получаю (при большой нагрузке на LAN) :

kernel BUG at drivers/net/macb.c:428!
Unable to handle kernel NULL pointer dereference at virtual address 00000000
pgd = c3a24000
[00000000] *pgd=239d3031, *pte=00000000, *ppte=00000000
Internal error: Oops: 817 [#1] PREEMPT
last sysfs file:
Modules linked in: vin_mod k16_mod scsi_wait_scan
CPU: 0 Not tainted (2.6.33 #33)
PC is at __bug+0x1c/0x28
LR is at __bug+0x18/0x28
pc : [<c002c8a4>] lr : [<c002c8a0>] psr: 60000013
sp : c39dbe50 ip : fefff200 fp : 00000040
r10: 00000100 r9 : c3884350 r8 : 000001a3
r7 : 000000d6 r6 : c3884320 r5 : c38435c0 r4 : 000001a2
r3 : 00000000 r2 : 00000102 r1 : 00001ab5 r0 : 0000002c
Flags: nZCv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user
Control: 0005317f Table: 23a24000 DAC: 00000015
Process elgatos (pid: 504, stack limit = 0xc39da270)
Stack: (0xc39dbe50 to 0xc39dc000)
be40: 000001a1 c01b0bf0 00000040 00000000
be60: c39dbf74 c3884350 00000040 0000000c 0000012c c037f55c c037f56c 002753e2
be80: c03941e8 c01fca7c c39dbdb8 00000102 c39da000 0000000c 00000001 00000003
bea0: 0000000a c038b460 00000000 c0049cb4 c39da000 00000000 00000015 00000015
bec0: 00000000 00000015 00000001 00000000 c39da000 41977216 41977434 c004a004
bee0: 00000015 c0028080 0000c350 ffffffff fefff000 c0028ae8 c482f000 00000032
bf00: 000001ff 000000fd 00001090 0000001c c4841080 00000008 00000208 c39da000
bf20: 41977216 41977434 c4841090 c39dbf40 000001e3 bf0065b8 20000013 ffffffff
bf40: 165ac476 00000001 00000000 c39d83c0 41977216 c39dbf88 00000003 00000208
bf60: 00000000 c0098be4 00000000 00000000 00000000 00000000 c39d83c0 00000003
bf80: c0029088 c0098ff4 00000000 00000000 00000000 00000001 00000002 00000004
bfa0: 40041000 c0028ee0 00000002 00000004 0000000c 41977216 00000208 41977216
bfc0: 00000002 00000004 40041000 00000003 40041000 003d0f00 40029864 41977434
bfe0: 00000000 41976c00 40033624 40033ca4 80000010 0000000c 00000000 00000000
[<c002c8a4>] (__bug+0x1c/0x28) from [<c01b0bf0>] (macb_poll+0x164/0x248)
[<c01b0bf0>] (macb_poll+0x164/0x248) from [<c01fca7c>] (net_rx_action+0x68/0x180)
[<c01fca7c>] (net_rx_action+0x68/0x180) from [<c0049cb4>] (__do_softirq+0x78/0x100)
[<c0049cb4>] (__do_softirq+0x78/0x100) from [<c004a004>] (irq_exit+0x44/0x64)
[<c004a004>] (irq_exit+0x44/0x64) from [<c0028080>] (asm_do_IRQ+0x80/0xa0)
[<c0028080>] (asm_do_IRQ+0x80/0xa0) from [<c0028ae8>] (__irq_svc+0x48/0x8c)
Exception stack(0xc39dbef8 to 0xc39dbf40)
bee0: c482f000 00000032
bf00: 000001ff 000000fd 00001090 0000001c c4841080 00000008 00000208 c39da000
bf20: 41977216 41977434 c4841090 c39dbf40 000001e3 bf0065b8 20000013 ffffffff
[<c0028ae8>] (__irq_svc+0x48/0x8c) from [<bf0065b8>] (rchan_read+0x324/0x5b0 [k16_mod])
[<bf0065b8>] (rchan_read+0x324/0x5b0 [k16_mod]) from [<c0098be4>] (vfs_read+0xac/0x158)
[<c0098be4>] (vfs_read+0xac/0x158) from [<c0098ff4>] (sys_read+0x3c/0x68)
[<c0098ff4>] (sys_read+0x3c/0x68) from [<c0028ee0>] (ret_fast_syscall+0x0/0x28)
Code: e1a01000 e59f000c eb006336 e3a03000 (e5833000)
---[ end trace 4ca592b1fafad532 ]---
Kernel panic - not syncing: Fatal exception in interrupt
[<c002e998>] (unwind_backtrace+0x0/0xdc) from [<c0044878>] (panic+0x58/0x12c)
[<c0044878>] (panic+0x58/0x12c) from [<c002cd48>] (die+0x16c/0x198)
[<c002cd48>] (die+0x16c/0x198) from [<c002f6c0>] (__do_kernel_fault+0x68/0x80)
[<c002f6c0>] (__do_kernel_fault+0x68/0x80) from [<c002f89c>] (do_page_fault+0x1c4/0x1dc)
[<c002f89c>] (do_page_fault+0x1c4/0x1dc) from [<c0028318>] (do_DataAbort+0x38/0x9c)
[<c0028318>] (do_DataAbort+0x38/0x9c) from [<c0028a8c>] (__dabt_svc+0x4c/0x60)
Exception stack(0xc39dbe08 to 0xc39dbe50)
be00: 0000002c 00001ab5 00000102 00000000 000001a2 c38435c0
be20: c3884320 000000d6 000001a3 c3884350 00000100 00000040 fefff200 c39dbe50
be40: c002c8a0 c002c8a4 60000013 ffffffff
[<c0028a8c>] (__dabt_svc+0x4c/0x60) from [<c002c8a4>] (__bug+0x1c/0x28)
[<c002c8a4>] (__bug+0x1c/0x28) from [<c01b0bf0>] (macb_poll+0x164/0x248)
[<c01b0bf0>] (macb_poll+0x164/0x248) from [<c01fca7c>] (net_rx_action+0x68/0x180)
[<c01fca7c>] (net_rx_action+0x68/0x180) from [<c0049cb4>] (__do_softirq+0x78/0x100)
[<c0049cb4>] (__do_softirq+0x78/0x100) from [<c004a004>] (irq_exit+0x44/0x64)
[<c004a004>] (irq_exit+0x44/0x64) from [<c0028080>] (asm_do_IRQ+0x80/0xa0)

[<c0028080>] (asm_do_IRQ+0x80/0xa0) from [<c0028ae8>] (__irq_svc+0x48/0x8c)
Exception stack(0xc39dbef8 to 0xc39dbf40)
bee0: c482f000 00000032
bf00: 000001ff 000000fd 00001090 0000001c c4841080 00000008 00000208 c39da000
bf20: 41977216 41977434 c4841090 c39dbf40 000001e3 bf0065b8 20000013 ffffffff
[<c0028ae8>] (__irq_svc+0x48/0x8c) from [<bf0065b8>] (rchan_read+0x324/0x5b0 [k16_mod])
[<bf0065b8>] (rchan_read+0x324/0x5b0 [k16_mod]) from [<c0098be4>] (vfs_read+0xac/0x158)
[<c0098be4>] (vfs_read+0xac/0x158) from [<c0098ff4>] (sys_read+0x3c/0x68)
[<c0098ff4>] (sys_read+0x3c/0x68) from [<c0028ee0>] (ret_fast_syscall+0x0/0x28)


Причем строка 428 в исходнике macb.c выглядит так

BUG_ON(frag != last_frag);

кто-то же встречал такую беду ? отзовитесь пожалуйста