https://sourceware.org/bugzilla/show_bug.cgi?id=34821

            Bug ID: 34821
           Summary: GNU ld MIPS dereferences a NULL hash entry while
                    asserting a MINUS_ONE TLS value
           Product: binutils
           Version: 2.46
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: ld
          Assignee: unassigned at sourceware dot org
          Reporter: basavesh.shivakumar at wsu dot edu
  Target Milestone: ---

Created attachment 17141
  --> https://sourceware.org/bugzilla/attachment.cgi?id=17141&action=edit
repro

I'm reporting this bug on behalf of my team: Ananta Srikar Puranam, Emily West,
Dr. Swarn Priya, Dr. Peng Gao and Dr. Basavesh Ammanaghatta Shivakumar.

Summary
=======

An O32 R_MIPS_TLS_GOTTPREL relocation against symbol zero whose st_value is the
MINUS_ONE sentinel makes released GNU ld segfault in final links and current
commit 1785516 evaluate h->root through NULL inside the assertion at
elfxx-mips.c:3567; the exact one-byte 0xfffffffe control and the conforming
zero-valued symbol control pass static and PIE links.

Reproduction
============

1. Download and extract
`bfd-mips-tls-gottprel-minus-one-null-hash-entry-reproducer.tar.gz`.
2. Run the following command with an ld.bfd or all-target `ld-new` binary that
   supports the emulation used by the reproducer:

       BFD=/path/to/ld-new sh repro.sh

The archive contains the fixed input files, independent controls, helper
scripts needed by `repro.sh`, and their checksums. The script uses a temporary
directory and does not modify the source tree.

Observed result
===============

An O32 R_MIPS_TLS_GOTTPREL relocation against symbol zero whose st_value is the
MINUS_ONE sentinel makes released GNU ld segfault in final links and current
commit 1785516 evaluate h->root through NULL inside the assertion at
elfxx-mips.c:3567; the exact one-byte 0xfffffffe control and the conforming
zero-valued symbol control pass static and PIE links.

Expected result
===============

GNU ld should either produce the correct output or reject the input with a
normal diagnostic. It should not crash, hang, execute undefined behavior, read
outside an input-derived allocation, or produce the described malformed
output.

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

Reply via email to