Re: NIC fails in RELEASE build CE 5.0

Tech-Archive recommends: Repair Windows Errors & Optimize Windows Performance

It Works!! I added a :

"SysIntr"=dword:1A
in the NE20001 registry entry to force it to use the already existing
irq/Sysintr entry.

Thanks Paul!! I owe you a beer!


"Paul G. Tobey [eMVP]" <p space tobey no spam AT no instrument no spam DOT
com> wrote in message news:O7zcXWjoHHA.1388@xxxxxxxxxxxxxxxxxxxxxxx
If the SYSINTR values don't match, your driver's interrupt processing
won't work.

When the driver called InterruptInitialize(), the hardware interrupt
should be enabled by the OAL or bus driver code.

Paul T.

"allanL" <alunk AT MCWW COM> wrote in message
news:OxV2jphoHHA.3460@xxxxxxxxxxxxxxxxxxxxxxx
Thanks Paul!
Basically, I have set the NIC to use IRQ 0x0A, and I have left it as
Either PCI or ISA in the BIOS (tried it both ways, and I am not really
sure what the difference is) I traced through the interrupt
initialization code, and it looks like the driver just takes the IRQ and
maps it to the first unused SYSINTR slot, then it remaps that SYSINTR to
the irq. This looks OK, though the IRQ is already mapped to a different
SYSINTR earlier in the code.

What I don't see is where the hardware interrupt is actually enabled. My
hunch is that is left dissabled. Any idea where this is done in the CEPC
version? Do I have access to the code?

thanks again!


"Paul G. Tobey [eMVP]" <p space tobey no spam AT no instrument no spam
DOT com> wrote in message news:uLyFYShoHHA.596@xxxxxxxxxxxxxxxxxxxxxxx
Make sure that you understand what the interrupt settings are telling
the driver. Sometimes it's a physical interrupt number that means
something in the context of the bus (IRQ5 on the ISA bus, for example).
Sometimes it's selecting one of the several interrupt outputs that the
Ethernet controller can support (that is, my controller might have IRQa,
IRQb, IRQc, and IRQd outputs from the chip, expecting you to connect
those to the bus interrupt lines and the setting that you're changing in
the registry might be telling the driver which of the chip outputs to
use). The interrupt registry entry might also be a SYSINTR number which
has nothing to do with the bus interrupt number or anything in hardware
(read the help). The low-level Ethernet stuff for VMINI usually doesn't
use interrupts at all, so this is a reasonable explanation for why
things don't work for you when you're trying to use the real driver...

Paul T.

"AllanL" <allan AT ByteWiseSG D0T com> wrote in message
news:%23E$IkZFoHHA.1216@xxxxxxxxxxxxxxxxxxxxxxx
I have done more research, and I have found that I can set the physical
IRQ on the board to one value and tell the driver it is another, and I
get identical results as when I give the driver the correct value. I am
wondering if the card never gets interrupts (It works fine with EDBG)

Is there some other setting for IRQs. Initialization looks like it
works:
4294785328 PID:e1fac9be TID:e1fcbae6 0x81fce8c0: NE2000:CardRamTest -
RamBase 0x4000, RamEnd 0x8000
4294785445 PID:e1fac9be TID:e1fcbae6 0x81fce8c0: NE2000:RegisterAdapter
CardInitialize -- Success
4294785549 PID:e1fac9be TID:e1fcbae6 0x81fce8c0: NE2000:RegisterAdapter
Xmit Start (0x4000, 0x40) : Rcv Start (0x4c00, 0x4c) : Rcv End (0x8000,
0x80)
4294785640 PID:e1fac9be TID:e1fcbae6 0x81fce8c0: NE2000:
PermanentAddress [ 00-01-45-00-1b-22 ]
4294785711 PID:e1fac9be TID:e1fcbae6 0x81fce8c0: NE2000:RegisterAdapter
CardSetup
4294785720 PID:e1fac9be TID:e1fcbae6 0x81fce8c0: NE2000:RegisterAdapter
CardSetup -- Success
4294785879 PID:e1fac9be TID:e1fcbae6 0x81fce8c0: ***NDIS*** (a, 534)
4294785887 PID:e1fac9be TID:e1fcbae6 0x81fce8c0:
==>NdisMRegisterInterrupt: Miniport 00032FB8 Vector 10
4294785900 PID:e1fac9be TID:e1fcbae6 0x81fce8c0: NDIS::
CeConnectInterrupt() --> IRQ[10] SYSINTR[16]
4294785911 PID:e1fac9be TID:e1fcbae6 0x81fce8c0: ***NDIS*** (a, 753)
4294785920 PID:e1fac9be TID:e1fcbae6 0x81fce8c0:
<==NdisMRegisterInterrupt: Miniport 00032FB8, Status 0
4294786003 PID:e1fac9be TID:e1fcbae6 0x81fce8c0: NE2000:RegisterAdapter
Interrupt Connected
4294786013 PID:e1fac9be TID:e1fcbae6 0x81fce8c0:
==>NdisMRegisterAdapterShutdownHandler: Miniport 00032FB8
4294786025 PID:e1fac9be TID:e1fcbae6 0x81fce8c0:
<==NdisMRegisterAdapterShutdownHandler: Miniport 00032FB8
4294786036 PID:e1fac9be TID:e1fcbae6 0x81fce8c0: NE2000:RegisterAdapter
OK
4294786045 PID:e1fac9be TID:e1fcbae6 0x81fce8c0:
==>NdisCloseConfiguration: ConfigurationHandle 000339A0
4294786057 PID:e1fac9be TID:e1fcbae6 0x81fce8c0:
<==NdisCloseConfiguration: ConfigurationHandle 000339A0
4294786538 PID:e1fac9be TID:e1fcbae6 0x81fce8c0: NE2000:Initialize
succeeded
4294786572 PID:e1fac9be TID:e1fcbae6 0x81fce8c0:
ndisMInitializeAdapter: Miniport 00032FB8, InitializeHandler returned 0
4294786585 PID:e1fac9be TID:e1fcbae6 0x81fce8c0: ***NDIS*** (c, 712)
4294786593 PID:e1fac9be TID:e1fcbae6 0x81fce8c0: ==>ndisMDoRequests
4294786601 PID:e1fac9be TID:e1fcbae6 0x81fce8c0: ***NDIS*** (c, 739)



After everything is loaded, I get this sequence repeated while DHCP
keeps trying..........Note the timeout in DHCP....
.
.
.
.
3020051 PID:e1fac9be TID:61f40fd2 0x81f40cc0: ==>ndisMProcessDeferred
3020072 PID:e1fac9be TID:61f40fd2 0x81f40cc0: <==ndisMProcessDeferred
3022045 PID:e1fac9be TID:61f40fd2 0x81f40cc0: ==>ndisMProcessDeferred
3022054 PID:e1fac9be TID:61f40fd2 0x81f40cc0: <==ndisMProcessDeferred
3022566 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: DHCP:+ProcessAutoIP
3022574 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: DHCP: +DhcpInitSock:
pDhcp 0X30320
3022584 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0:
+WSHOpenSocket(0x85E44,0x85E48,0x85E50,0x813F8CC,0x85F38,0x85F3C)
3022596 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: -WSHOpenSocket Return 0
3022605 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: DHCP: -DhcpInitSock:
pDhcp 0x30320/0x85E30 Ret: 0
3022634 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: +BuildDhcpPkt: DISCOVER
pDhcp 0x30320/0x85E30
3022645 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: DHCP: TX DISCOVER for
NE20001
3022655 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: -BuildDhcpPkt: pDhcp
0x30320/0x85E30 Len 274
3022665 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: +SendDhcpPkt: pDhcp 30320
pPkt 813F9EC
3022851 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: *GetIPPacket: Packet
3768C Pool 37600
3022861 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ***NDIS*** (17, 694)
3022869 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ==>ndisMSend
3022877 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ***NDIS*** (b, 2912)
3022886 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ==>ndisMQueueWorkItem
3022894 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ***NDIS*** (b, 2945)
3022961 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: <==ndisMQueueWorkItem
3022969 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ==>ndisMProcessDeferred
3022978 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ***NDIS*** (17, 1412)
3022986 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ==>ndisMStartSends
3022995 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ***NDIS*** (17, 1455)
3023003 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: Sending packet 0x0003768C
3023012 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ***NDIS*** (17, 1022)
3023021 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ==>ndisMSendComplete
3023029 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ***NDIS*** (17, 1024)
3023038 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: packet 0x3768c
3023048 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ***NDIS*** (17, 57)
3023056 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: Completed 0x0
3023064 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: FreeIPPacket (3): Packet
3768C Pool 37600
3023074 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: PplFree: Addr 40580
3023083 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ***NDIS*** (17, 1056)
3023119 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: <==ndisMSendComplete
3023128 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ***NDIS*** (17, 1470)
3023136 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: Complete is pending
3023145 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ***NDIS*** (17, 1491)
3023153 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: <==ndisMStartSends
3023162 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: <==ndisMProcessDeferred
3023172 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: ***NDIS*** (17, 760)
3023180 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: <==ndisMSend
3023188 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: *DhcpSendDown: Dest
67/375ffea9, Sent 300/300 Status 0
3024003 PID:e1fac9be TID:61f40fd2 0x81f40cc0: ==>ndisMProcessDeferred
3024011 PID:e1fac9be TID:61f40fd2 0x81f40cc0: <==ndisMProcessDeferred

3024200 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: SendDhcpPkt: Select()
timed out $$$$$$$$$$$$$$$$$$$$$

3024229 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: -SendDhcpPkt: pDhcp 30320
pPkt 813F9EC Ret: 1
3024240 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0:
+WSHNotify(0x86040,0x85E30,0x0,0x0,128)
3024250 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: WSHNotify got a
NOTIFY_CLOSE freeing Socket Context 86040
3024262 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: -WSHNotify Return 0
3024270 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: DHCP:StartAutoIPTimer -
AutoIP event scheduled
3024282 PID:e1fac9be TID:c1ec1c7e 0x81efc2a0: DHCP:-ProcessAutoIP
3025985 PID:e1fac9be TID:61f40fd2 0x81f40cc0: ==>ndisMProcessDeferred
3025995 PID:e1fac9be TID:61f40fd2 0x81f40cc0: <==ndisMProcessDeferred
3027968 PID:e1fac9be TID:61f40fd2 0x81f40cc0: ==>ndisMProcessDeferred
.
.

Thanks for any help.

AL


"Paul G. Tobey [eMVP]" <p space tobey no spam AT no instrument no spam
DOT com> wrote in message
news:%23oeKX8wnHHA.1244@xxxxxxxxxxxxxxxxxxxxxxx
There's a "how to use vmini" document in the VMINI directory. Make
sure that you don't have the VMINI entries in the registry. You
should be able to leave the DLL in the image, if you want.

Paul T.

"allanL" <alunk AT MCWW COM> wrote in message
news:eQ9pWpwnHHA.4192@xxxxxxxxxxxxxxxxxxxxxxx
I think so....
How exactly do I remove it? I used
BSP_NOSHAREETH=1
KERNELNOSHAREETH=1
and un-checked (IMGNOKITL=1) in the build options. Is there
something else I need to do? I don't see vmini loading when I use
SerialDebug and look at the trace. How can I tell what exactly is
included in my build? 4.2 had a batch file I could modify/read to see
exactly what build variables were set. I can't seem to find it in
5.0.

thx



"Paul G. Tobey [eMVP]" <p space tobey no spam AT no instrument no
spam DOT com> wrote in message
news:O$R6qeunHHA.3656@xxxxxxxxxxxxxxxxxxxxxxx
And you did remove VMINI in the release build, right? The Ethernet
controller can't serve two masters, so, if your OAL code is trying
to operate it for VMINI/KITL and you also are trying to run the
right driver on top of it, it won't work.

Paul T.

"AllanL" <allan AT ByteWiseSG D0T com> wrote in message
news:%23HSF3nnnHHA.3512@xxxxxxxxxxxxxxxxxxxxxxx
I have a CEPC-GEODE PC/104 board with an NE2000 ISA NIC. When I use
the VMINI driver in a debug build it works fine. In the release
build, DHCP fails. Searching, it seems that similar problems are
form not removing the vmini driver and KITL formthe build. I have
done this--at least I think I have.

Static IP also doesn't work--the board won't respond to a ping.

Here is what I have done in the build:

IMGNOKITL=1 (UN-check box - Build Options)
KERNELNOSHAREETH=1
BSP_NOSHAREETH=1
BSP_NIC_NE2000_ISA=1

I also have:

BSP_NOPCIBUS=1
BSP_NOPCMCIA=1 (These are also in the debug build)


In the DEBUG Build, none of these are set.
In the registry, I hav left the default settings, with the
exception of
setting port (0x2c0) and IRQ (5) as follows:
[HKEY_LOCAL_MACHINE\Comm\NE2000]
"DisplayName"=LOC_DISPLAYNAME_COMM_NE2000
"Group"="NDIS"
"ImagePath"="ne2000.dll"

[HKEY_LOCAL_MACHINE\Comm\NE2000\Linkage]
"Route"=multi_sz:"NE20001"

[HKEY_LOCAL_MACHINE\Comm\NE20001]
"DisplayName"=LOC_DISPLAYNAME_COMM_NE20001
"Group"="NDIS"
"ImagePath"="ne2000.dll"

[HKEY_LOCAL_MACHINE\Comm\NE20001\Parms]
"BusNumber"=dword:0
"BusType"=dword:1
"InterruptNumber"=dword:05
"IoBaseAddress"=dword:2C0
"Transceiver"=dword:3
"CardType"=dword:0

[HKEY_LOCAL_MACHINE\Comm\NE20001\Parms\TcpIp]
"EnableDHCP"=dword:1
"DefaultGateway"=""
"UseZeroBroadcast"=dword:0
"IpAddress"="0.0.0.0"
"Subnetmask"="0.0.0.0"

The CE board sends a "DHCP Discover message", and it gets an Offer
message
from the DHCP server, but the message is ignored--I never see a
DHCP
Request--instead it sends several more "DHCP Discover" messages.

I enabled serialDebug, but kept :
KERNELNOSHAREETH=1
BSP_NOSHAREETH=1
I get debug message through the serial port and I can see that the
NIC is initialized and DHCP loads and tries to DHCP, but never sees
the replys from the DHCP server (They are sent)

It looks like the board cannot receive any data.

I am really stuck on this one!

thx

















.

Relevant Pages

  • Re: wireless card driver problem about the "done" bit in the descriptor && the bail caused b
    ... the VxWorks, ... or a "rxdp bail", the whileloop break, while no interrupt to make ... it off and do a jobQueuePostto wake up tNet0. ... processing driver interrupt events in a timely manner. ...
    (comp.os.vxworks)
  • Re: 2.6.19 and up to 2.6.20-rc2 Ethernet problems x86_64
    ... If the system fails in ACPI mode, but works in non-ACPI mode, ... Using ACPI for SMP configuration information ... Using IOAPIC for interrupt routing ... Non-volatile memory driver v1.2 ...
    (Linux-Kernel)
  • Re: [PATCH] uio: User IRQ Mode
    ... This patch adds a "User IRQ Mode" to UIO. ... In this mode the user space driver ... is responsible for acknowledging and re-enabling the interrupt. ...
    (Linux-Kernel)
  • [PATCH 06/13] Char: cyclades, overall cleanup
    ... * Prevent users from opening non-existing Z ports. ... * Implemented full support for non-standard baud rates. ... * Driver now reports physical instead of virtual memory addresses. ... * Fixed SMP locking in Cyclom-Y interrupt handler. ...
    (Linux-Kernel)
  • [PATCH 13/79] cyclades: overall cleanup
    ... * Prevent users from opening non-existing Z ports. ... * Implemented full support for non-standard baud rates. ... * Driver now reports physical instead of virtual memory addresses. ... * Fixed SMP locking in Cyclom-Y interrupt handler. ...
    (Linux-Kernel)