Bug 1771 - ld.so segfaults on multiGOT binaries with >16k global symbols produced by binutils on MIPS
Summary: ld.so segfaults on multiGOT binaries with >16k global symbols produced by bin...
Status: RESOLVED INVALID
Alias: None
Product: glibc
Classification: Unclassified
Component: ports (show other bugs)
Version: unspecified
: P2 normal
Target Milestone: ---
Assignee: drow@false.org
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2005-10-31 19:02 UTC by Nathanael Nerode
Modified: 2016-05-20 19:50 UTC (History)
2 users (show)

See Also:
Host: mips*-*-linux*
Target: mips*-*-linux*
Build: mips*-*-linux*
Last reconfirmed:
fweimer: security-


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Nathanael Nerode 2005-10-31 19:02:18 UTC
I'm not sure which component the dynamic linker goes with.    
    
In any case, this was discussed at length in Debian, terminating in the    
message http://lists.debian.org/debian-mips/2005/10/msg00031.html by   
Thiemo Seufer:   
> > * or a bug in ld.so -- inability to handle correctly specified multiple  
GOTs  
> > for more than 16k global symbols  
> 
> That (it shouldn't segfault), and/or potentially also a bug in ld which  
> leads to failure for large MultiGOT binaries. 
 
In the spirit of actually reporting bugs, rather than keeping them known only 
to a secret cabal, I'm reporting this. 
 
The only current known testcase is libgcj from GCC.  It's probably possible to 
artificially construct MultiGOT binaries with >16K of global symbols, however.
Comment 1 Roland McGrath 2006-02-21 02:14:20 UTC
For the trunk, mips has moved to the ports repository.
Daniel should be able to commit his fixes there directly.
Items needing merge in the 2.3 branch can be reassigned to me.
Comment 2 Daniel Jacobowitz 2006-06-08 15:58:25 UTC
Thiemo, is the conclusion that this was the same as the ld bug David Daney just
tracked down?
Comment 3 Thiemo Seufer 2006-06-08 16:15:50 UTC
Yes.
Comment 4 Daniel Jacobowitz 2006-06-08 17:18:40 UTC
Not a glibc bug, then.  I'll follow the binutils bug.