On 4/7/09 8:33 AM, Lassi A.Tuura wrote:
> At the end of my build, I do "make check_both", and I get the output
> listed below. Is this a known problem, or is this a feature of my
> build? Either way, any ideas where the problem may be and how to fix it?
Getting big numbers correct in lazytypes has been really hard, partly
because it doesn't know the word width of the underlying GCC. This error is
very minor and probably won't cause you any problems, so you can ignore it.
> The two numbers are 2^32-1 and 2^64-1. I assume the bit field it
> complains about is in types.cc:
>
> struct Boo {
> unsigned int i:31;
> } foo;
You are probably correct.
> I am not entirely sure how the max value of that could be either 2^32-1
> or 2^64-1.
Note that what you're comparing is
root.type.variantOf.members.0.type.members.0.type.bitfieldOf.max.value
so the test is actually asking for the max value of 'unsigned int', which
indeed could be 2^32-1 (if unsigned int is a 4-byte number) or 2^64-1 (if
unsigned int is an 8-byte number).
> PS. I know between little and nothing about mercurial; is it possible to
> check out a "tag" or some fixed labelled revision of dehydra stuff somehow?
Each revision has a changeset ID: see the strings at
http://hg.mozilla.org/users/tglek_mozilla.com/dehydra-gcc/ such as
343ab385b885 and 87a11f4441ae
To find out what revision you currently have, use `hg log -r.`
It's also possible to have named tags, but the last named tag was
'dehydra-0.9' which was quite a while back.
--BDS
_______________________________________________
dev-static-analysis mailing list
[email protected]
https://lists.mozilla.org/listinfo/dev-static-analysis
