On Wed, Jan 16, 2013 at 12:02 PM, Dave Reisner <[email protected]> wrote: > Trust this value, instead of injecting our own, as this could be > something as simple as ENOENT, rather than an alarming ENOMEM. > --- > libkmod/libkmod.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/libkmod/libkmod.c b/libkmod/libkmod.c > index b3e1d6b..0d9cd80 100644 > --- a/libkmod/libkmod.c > +++ b/libkmod/libkmod.c > @@ -816,7 +816,7 @@ KMOD_EXPORT int kmod_load_resources(struct kmod_ctx *ctx) > > fail: > kmod_unload_resources(ctx); > - return -ENOMEM; > + return -errno; > }
Does this actually work? if index_mm_open() fails it may had passed through munmap(), close() and free() before returning to kmod_load_resources(). The first 2 can touch errno. Worse than that, we also "goto fail" in index_mm_open() if the checks for magic or version failed, in which case errno will not be set and we will incorrectly return that this function succeeded. return -ENOMEM there maybe is not a good choice, but changing it to may actually introduce bugs. Maybe -EINVAL or -1 would be a better choice. Thoughts? Lucas De Marchi -- To unsubscribe from this list: send the line "unsubscribe linux-modules" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
