I bought a new ASUS AT5IONT-I mini itx board which works great, but the network never works after it's been initialised either via reboot or sleep , unless you pull the power cord and cold boot it again,
All the forums etc.. say use the r8168 driver from Realtek. Now there lies the problem, the source code for this driver is so similar to r8169 in the kernel and it's only supplied as a module, that isn't even a problem I worked out how to get it to build into the kernel itself under drivers/net/r8168, and got the kernel to build, I also figured out the gar system enough to get it to patch it and use my kernel config from ~/.minimyth
I will provide an update on this when I get home and try booting this kernel bzImage, it it works then my suspend/resume problem should be fixed.
I was just thinking while writing this email the fix might be as simple as changing the PCI id's in r8169 to not conflict with r8168 and hence this r8168 driver can be permanently patched into mini-myth. The current situation would require me to build minimyth everytime just for this frontend.
If I can trace the kernel bug down which I have been unable to do so as everyone says use the r8168 driver! then fixing in the kernel r8169 would be better.
75b1 includes the 2.6.36 Linux kernel, which claims to include some fixes to the r8169 that fix problems with r8168 hardware.
Like you, I would prefer the upstream kernel to support my hardware. However, if the the 2.6.36 kernel does not solve the problem, then we will need to blacklist hardware IDs that work with the r8168 driver but not the r8169 driver.
Could you give 0.23.1-75b2 a try? I have patched the kernel to include the r8168 driver, including patching the r8169 driver so that it should not claim 0x8168 realtek devices.
I am at work now but looking at the gar of it it's exactly the same thing I did, Can I ask did you make the kernel patch yourself or did you find it on the net. Only asking because I couldn't find any information anywhere and just figured it out by looking at other network card drivers in the kernel source, I ended up with the same patch anyway.
I noticed in the changelog for 76b1 that the r8168 driver has been removed because some r8169 fixes went into the newer kernel. I just tested that version and it dosen't fix the problem with suspend / resume that the realtek sources fix.