[Bug libctf/30013] New: [2.40 regression] Assertion failed: one_type != two_type, file libctf/ctf-dedup.c, line 2342, function sort_output_mapping

[ro at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=30013

            Bug ID: 30013
           Summary: [2.40 regression] Assertion failed: one_type !=
                    two_type, file libctf/ctf-dedup.c, line 2342, function
                    sort_output_mapping
           Product: binutils
           Version: 2.40
            Status: NEW
          Severity: normal
          Priority: P2
         Component: libctf
          Assignee: unassigned at sourceware dot org
          Reporter: ro at gcc dot gnu.org
                CC: nick.alcock at oracle dot com
  Target Milestone: ---

When I recently upgraded my self-compiled version used for gcc bootstraps from
2.39 to 2.40, a couple of testsuite regressions occured:

+FAIL: gcc.dg/debug/20020220-1.c -gctf (test for excess errors)

Excess errors:
Assertion failed: one_type != two_type, file
/vol/src/gnu/binutils/binutils-2.40/libctf/ctf-dedup.c, line 2342, function
sort_output_mapping
collect2: fatal error: ld terminated with signal 6 [Abort]

+FAIL: gcc.dg/debug/20020220-1.c -gctf execution test
+FAIL: gcc.dg/debug/20050907-1.c -gctf (test for excess errors)
+FAIL: gcc.dg/debug/pr36690-3.c -gctf (test for excess errors)
+FAIL: gcc.dg/debug/pr36690-3.c -gctf execution test
+FAIL: gcc.dg/debug/pr37616.c -gctf (test for excess errors)
+FAIL: gcc.dg/debug/pr37616.c -gctf execution test
+FAIL: gcc.dg/debug/pr41893-1.c -gctf (test for excess errors)
+FAIL: gcc.dg/debug/pr49032.c -gctf (test for excess errors)
+FAIL: gcc.dg/debug/pr65771.c -gctf (test for excess errors)

They happen on Solaris/sparc and x86, 32 and 64-bit.

It turns out that the assertion failure occurs when qsort_r is missing from
libc
(as on Solaris 11.3, unlike Solaris 11.4).

However, there's nothing Solaris-specific here: if I disable qsort_r in
config.status on Linux/i686

--- config.status.dist  2023-01-17 16:51:27.241499000 +0100
+++ config.status       2023-01-17 16:51:42.380104000 +0100
@@ -820,2 +820,2 @@
-S["NEED_CTF_QSORT_R_FALSE"]=""
-S["NEED_CTF_QSORT_R_TRUE"]="#"
+S["NEED_CTF_QSORT_R_FALSE"]="#"
+S["NEED_CTF_QSORT_R_TRUE"]=""
@@ -1050,2 +1049,0 @@
-D["HAVE_QSORT_R"]=" 1"
-D["HAVE_QSORT_R_ARG_LAST"]=" 1"

(there sems to be no way to do this otherwise, e.g. by presetting autoconf
cache variables), I get exactly the same failure, so there seems to be
something
wrong with ctf-qsort_r.c.

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