[Bug libdw/26773] sleb128 values near INT64_MAX/MIN not correctly read

[mark at klomp dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=26773

--- Comment #3 from Mark Wielaard <mark at klomp dot org> ---
(In reply to Tom Tromey from comment #1)
> I looked at this a little today.
> 
> In addition to this bug, I think the _unchecked variants
> have another bug; namely they do:
> 
>   const size_t max = len_leb128 (int64_t) - 1;
> 
> This limits the number of bytes read-- but it seems like
> it maybe ought to read until the first byte without the
> high bit set.

Yes, that is the bug (plus the same -1 we do inside __libdw_max_len_sleb128 for
the "checked" variant.

> FWIW gdb seems to rely on implementation-defined behavior
> here.  It does all the sleb work in an unsigned type
> and then casts it to signed on return.

That would probably work. Is the implementation-defined part that it depends on
signed numbers being represented as Two's complement? In that case I think we
are fine with that as fix.

-- 
You are receiving this mail because:
You are on the CC list for the bug.