Skip to contentSkip to conversation

BUG: LDP - smc911x region too small

DISCUSSION Russell King - ARM Linux 3 participants 6 messages Jan 19, 2009 – Feb 3, 2009

lore
Share & tools
lore
#0 lore raw

At the moment:

        ldp_smc911x_resources[0].start = cs_mem_base + 0x0;
        ldp_smc911x_resources[0].end   = cs_mem_base + 0xf;

However, the SMC911x driver wants to claim 256 bytes from this region. Indeed, its register set definitions appears to indicate that its needs more than 16 bytes of address space.

So, the question is what is the right fix?

Unfortunately, fixing this will cause problems with the SMSC911x patches, so I'd like to get an answer on this sooner rather than later please.

#1 lore raw
Quoted 5 lines · Russell King - ARM Linux

At the moment:

        ldp_smc911x_resources[0].start = cs_mem_base + 0x0;
        ldp_smc911x_resources[0].end   = cs_mem_base + 0xf;
Quoted 8 lines

However, the SMC911x driver wants to claim 256 bytes from this region. Indeed, its register set definitions appears to indicate that its needs more than 16 bytes of address space.

So, the question is what is the right fix?

Unfortunately, fixing this will cause problems with the SMSC911x patches, so I'd like to get an answer on this sooner rather than later please.

Here should be 0x100. Fixing this won't cause problems with smsc911x patches.

Stanley.

Quoted 4 lines

--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#2 lore raw
Quoted 10 lines · stanley.miao
Quoted 9 lines · Russell King - ARM Linux

At the moment:

        ldp_smc911x_resources[0].start = cs_mem_base + 0x0;
        ldp_smc911x_resources[0].end   = cs_mem_base + 0xf;

However, the SMC911x driver wants to claim 256 bytes from this region. Indeed, its register set definitions appears to indicate that its needs more than 16 bytes of address space.

Quoted 7 lines
Quoted 4 lines

So, the question is what is the right fix?

Unfortunately, fixing this will cause problems with the SMSC911x patches, so I'd like to get an answer on this sooner rather than later please.

Here should be 0x100. Fixing this won't cause problems with smsc911x patches.

Yes it will - if it gets fixed in the -rc series for the smc911x driver your patches will conflict.

There's more which needs fixing though - no platform data is being passed so the smc911x driver still fails to initialise.

#3 lore raw
Quoted 8 lines
Quoted 2 lines

Here should be 0x100. Fixing this won't cause problems with smsc911x patches.

Yes it will - if it gets fixed in the -rc series for the smc911x driver your patches will conflict.

There's more which needs fixing though - no platform data is being passed so the smc911x driver still fails to initialise.

Data sheet shows addressable range as 0x50 from base --> 0xFC. So 0xFF should be good.

I tried to send Russell data sheet but his emails is rejecting it.

Regards,
Richard W.

#4 lore raw
Quoted 10 lines · Richard
Quoted 9 lines
Quoted 3 lines

Here should be 0x100. Fixing this won't cause problems with smsc911x patches.

Yes it will - if it gets fixed in the -rc series for the smc911x driver your patches will conflict.

There's more which needs fixing though - no platform data is being passed so the smc911x driver still fails to initialise.

Quoted 8 lines

Data sheet shows addressable range as 0x50 from base --> 0xFC. So 0xFF should be good.

I tried to send Russell data sheet but his emails is rejecting it.

Regards,
Richard W.

It has been fixed.

http://marc.info/?l=linux-arm-kernel&m=123307584619895&w=2

Stanley.

#5 lore raw
Quoted 14 lines · stanley.miao
Quoted 13 lines · Richard
Quoted 9 lines
Quoted 3 lines

Here should be 0x100. Fixing this won't cause problems with smsc911x patches.

Yes it will - if it gets fixed in the -rc series for the smc911x driver your patches will conflict.

There's more which needs fixing though - no platform data is being passed so the smc911x driver still fails to initialise.

Data sheet shows addressable range as 0x50 from base --> 0xFC. So 0xFF should be good.

Quoted 7 lines
Quoted 6 lines

I tried to send Russell data sheet but his emails is rejecting it.

Regards,
Richard W.

It has been fixed.

No, it has not been fixed - it continues to fail in mainline kernels.

Quoted 1 line

http://marc.info/?l=linux-arm-kernel&m=123307584619895&w=2

That is a proposed patch for switching to the SMSC driver instead of the SMC driver. It's unclear at the moment whether it's acceptable to switch to the SMSC driver for the -rc kernels. Certainly just applying the patch above results in something which is also nonfunctional.

The motivation here is to _fix_ mainline kernels.