On 02/06/17 01:38, Luc Van Oostenryck wrote: > On Fri, Jun 2, 2017 at 2:30 AM, Ramsay Jones > <ramsay@xxxxxxxxxxxxxxxxxxxx> wrote: > >>> .SH OTHER OPTIONS >>> .TP >>> +.B \-fmemcpy-limit=COUNT >>> +By default, sparse will warn if \fBmemcpy()\fR (or \fBmemset()\fR, >>> +\fBcopy_from_user()\fR, copy_to_user()\fR) is called with a very large >>> +(known at compile-time) byte-count. COUNT is the value under which >>> +no such warning will be given. The default limit is 100000. >>> +. >>> +.TP >> >> So, in addition to -Wno-memcpy-max-count, you could turn the warning >> off with just -fmemcpy-limit=0. cool. >> >> Thanks! >> >> ATB, >> Ramsay Jones > > Well, for now setting the limit to 0 would just warn about any > non-zero memcpy/memset > but it's something that could very easily be added, sure. Yes, as I noted in another email, I didn't read the code correctly! (and in my patch I had a single -Wmem-limit=n argument which _did_ disable the check when n = 0). Naming issues aside (and I'm bad at naming, so don't listen to me on that), I like your -W[no-]memcpy-max-count and -fmemcpy-limit=n split. [You may want to remove the '=COUNT' from the commit message of the 2/3 patch]. Thanks again. ATB, Ramsay Jones -- To unsubscribe from this list: send the line "unsubscribe linux-sparse" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
