On 2014-04-08 17:02, Bjørn Mork wrote: > Felix Fietkau <n...@openwrt.org> writes: > >> I've seen this happen to other open source related projects using >> Marvell hardware as well, so the big question is whether Belkin can put >> enough pressure on them to get the source code released. >> >> Even if that happens, the source code will most likely need a rewrite or >> an insane amount of cleanup, as is typical for proprietary wifi drivers >> in the embedded space. >> >> There are many signs that if released, the source code to this driver is >> going to be horrible: weird function names, big module size, use of >> custom vendor-specific hostapd and wpa_supplicant drivers. This is most >> likely going to take a long time to resolve. > > I know these comments are based on experience, but I still feel you are > a bit too pessimistic here :-) I really do hope to be proven wrong on this one.
> After all, we do have the mwl8k driver in mainline and Marvell has > commited a lot to that, including the 8764 bits. It's not too unlikely > that they will add 8864 support as well, is it? I don't know how likely or unlikely it is. I also don't know how likely it is that they will care about the driver enough to keep AP mode support for embedded devices tested and working well, as opposed to just adding it as an afterthought. > And wrt the size: Some > of this is probably due to firmware being built into the module. And > some is debugging symbols. The rest is of course bloat mostly caused by > unnecessary reimplementation. Right. _______________________________________________ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
