On Thursday, August 19, 2010, Bojan Smojver wrote: > On Thu, 2010-08-19 at 01:02 +0200, Rafael J. Wysocki wrote: > > > I'm not a big fan of select in Kconfig. It ususally causes one to be surprised > > with the final choice of options in .config. > > > > Does that work at all if CRYPTO_ALGAPI is not set? > > My understanding is that these (LZO_COMPRESS/LZO_DECOMPRESS) do not > depend on anything. But, I can test, just to be sure. > > > I'd rather wouldn't like the above to be unconditional. If not for anything > > else, I think it would be nice to be able to switch the compression off for > > debugging purposes (like when we're not sure why the image is corrupted or > > similar). A kernel command line switch would suffice for that IMO. > > OK. I think the cleanest way may be to have a set of two "real" > save_image()/load_image() functions. One set with compression, the other > without, which are then called by save_image()/load_image(). OK? Sounds reasonable. > > > + if (!i) > > > break; > > > > This statement is profoundly hard to decode. Perhaps there's a better name for 'i'? > > Like 'size' or similar? > > Right. Maybe I'm a bit too economical with names. Usually things like > 'i' stand for a generic counter, which this is. I'll use something else > if you don't like it. > > > Also please make it possible to switch the compression off here too. > > > > The comments above seem to apply to the remaining code too. > > OK, I'll prepare a patch that has compression/decompression as optional, > using kernel command line switch. Obviously, if one attempts to thaw > from an image that is not in an expected state (compressed/not > compressed), this will wreak havoc. But, I guess that's understandable. Yes, it is. Thanks, Rafael _______________________________________________ linux-pm mailing list linux-pm@xxxxxxxxxxxxxxxxxxxxxxxxxx https://lists.linux-foundation.org/mailman/listinfo/linux-pm
