https://sourceware.org/bugzilla/show_bug.cgi?id=32260
--- Comment #10 from Jonny Weir ---
(In reply to Michael Matz from comment #9)
> So, I have two patches for you, see above. It would be nice to know that
> they work independendly, so if you could first test the first patch in
> isolation,
>
https://sourceware.org/bugzilla/show_bug.cgi?id=32260
--- Comment #4 from Jonny Weir ---
(In reply to Alan Modra from comment #3)
> Looks to me that merge.c:177 should be unsigned int newnb ...
> By calculating newnb as unsigned long we get to a count of 2^32 and overflow
> table->nbuckets withou
https://sourceware.org/bugzilla/show_bug.cgi?id=32260
--- Comment #2 from Jonny Weir ---
Ha, thanks :-)
Some additional logging would appear to indicate that the next estimation is
indeed quite aggressive:
XXX bfdtab->count=1713 table->nbuckets=524288
XXX bfdtab->count=1714 table->nbuckets=5242
: UNCONFIRMED
Severity: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: jonny.weir at clearpool dot io
Target Milestone: ---
Hi,
Just to note, this may be related to the following bug I raised previously that
was
https://sourceware.org/bugzilla/show_bug.cgi?id=31009
--- Comment #11 from Jonny Weir ---
Hi Michael,
(In reply to Michael Matz from comment #10)
> (In reply to Jonny Weir from comment #7)
> > I made the following change:
>
> Thanks!
>
> > XXX resize 1: count=1598 added=1086327410 newnb=104857
https://sourceware.org/bugzilla/show_bug.cgi?id=31009
--- Comment #9 from Jonny Weir ---
Hi Nick
(In reply to Nick Clifton from comment #8)
> Hi Jonny,
>
> (In reply to Jonny Weir from comment #3)
>
> > > Are you able to attach a debugger to the linker and discover the values
> > > that
> >
https://sourceware.org/bugzilla/show_bug.cgi?id=31009
--- Comment #7 from Jonny Weir ---
I made the following change:
diff --git a/bfd/merge.c b/bfd/merge.c
index f21154dcd45..3b4ccfb86df 100644
--- a/bfd/merge.c
+++ b/bfd/merge.c
@@ -175,26 +175,43 @@ sec_merge_maybe_resize (struct sec_merge_ha
https://sourceware.org/bugzilla/show_bug.cgi?id=31009
--- Comment #5 from Jonny Weir ---
Ignore that last message, it is misleading, this is a more accurate
representation of what is happening with the values:
bfdtab->count + 1 = 1 | table->nbuckets = 8192 | table->nbuckets * 2 / 3 = 5461
bfdtab
https://sourceware.org/bugzilla/show_bug.cgi?id=31009
--- Comment #4 from Jonny Weir ---
I was able to add some logging to get the values out and it seems that things
are overflowing at some point. Some example output:
bfdtab->count + 1 = 1 | table->nbuckets = 5461
bfdtab->count + 1 = 2 | tab
https://sourceware.org/bugzilla/show_bug.cgi?id=31009
--- Comment #3 from Jonny Weir ---
Hi Nick,
(In reply to Nick Clifton from comment #1)
> Hi Jonny,
>
> (In reply to Jonny Weir from comment #0)
>
> > linking stage when -O3 is used (-O2 builds and links correctly). To be
> > clear, the only
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: jonny.weir at clearpool dot io
Target Milestone: ---
Hi,
I am building a complex, large-scale C++23 project using gcc12 in Debian Linux
(sid) with kernel 6.5.0-1-amd64. When using binutils 2.39, the linking stage
11 matches
Mail list logo
