Forest Bond wrote: > > The following git repository has a vt6655 branch and a vt6656 branch: > > http://www.alittletooquiet.net/media/code/vt665x.git > > The vt6655 branch contains pretty much all of the work I'll likely be able to do > for that driver. The vt6656 branch needs more work. I've not gone farther than > build-testing either, at this point. It would be a little easier for looking at the code if both vt6655 and vt6656 directories were in the same branch of the git repo, but it's not a big deal. > As I've mentioned before, some of this kind of work is really a bit outside my > usual domain, so if you do look at it, I'd appreciate feedback. There are a number of warnings when compiling the vt6655 driver on x86_64 architecture that arise from pointers having 64 bits while integers have 32 bits. I think the warnings will not cause any problems. There is, however, one error, namely: CC [M] drivers/staging/vt6655/device_main.o drivers/staging/vt6655/device_main.c:347: error: initializer element is not constant drivers/staging/vt6655/device_main.c:347: error: (near initialization for ‘device_id_table[0].driver_data’) The following patch fixes the compilation by removing the integer cast on chip_info_table: Index: vt665x/drivers/staging/vt6655/device_main.c =================================================================== --- vt665x.orig/drivers/staging/vt6655/device_main.c +++ vt665x/drivers/staging/vt6655/device_main.c @@ -344,7 +344,7 @@ static CHIP_INFO chip_info_table[]= { }; static struct pci_device_id device_id_table[] __devinitdata = { -{ 0x1106, 0x3253, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (int)&chip_info_table[0]}, +{ 0x1106, 0x3253, PCI_ANY_ID, PCI_ANY_ID, 0, 0, &chip_info_table[0]}, { 0, } }; #endif ========================== Driver vt6656 also does not build, but that does not seem to be 64-bit related. If it builds on your system, let me know and I'll send you the error messages. I didn't look that thoroughly at the code, but I did notice one thing. The constructions like "#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20)" are in the code so that it can build on a number of different kernels. Once it is in staging, it only has to build on the current kernel and you can trim out those conditionals. Larry -- To unsubscribe from this list: send the line "unsubscribe linux-wireless" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
