[Bug symtab/17276] New: gdb needs to coalesce multiple line number entries for the same (non-prologue) line

["dje at google dot com" <sourceware-bugzilla at sourceware dot org>]

https://sourceware.org/bugzilla/show_bug.cgi?id=17276

            Bug ID: 17276
           Summary: gdb needs to coalesce multiple line number entries for
                    the same (non-prologue) line
           Product: gdb
           Version: HEAD
            Status: NEW
          Severity: normal
          Priority: P2
         Component: symtab
          Assignee: dje at google dot com
          Reporter: dje at google dot com

watchpoint-reuse-slot.exp fails with clang.
The reason is that the test does a stepi to step into the middle of a line and
expects gdb to print the pc value (which gdb will do when in the middle of a
line).

  for (i = 0; i < 10; ++i) continue;

clang emits several line number entries for that line,
each with a different discriminator.

The Directory Table:
  ../../../binutils-gdb/gdb/testsuite/gdb.base

 The File Name Table:
  Entry Dir     Time    Size    Name
  1     1       0       0       watchpoint-reuse-slot.c

 Line Number Statements:
  Extended opcode 2: set Address to 0x400600
  Advance Line by 29 to 30
  Copy
  Set column to 3
  Set prologue_end to true
  Special opcode 160: advance Address by 11 to 0x40060b and Line by 1 to 31
  Set column to 8
  Special opcode 106: advance Address by 7 to 0x400612 and Line by 3 to 34
  Extended opcode 4: set Discriminator to 4
  Special opcode 103: advance Address by 7 to 0x400619 and Line by 0 to 34
  Extended opcode 4: set Discriminator to 2
  Special opcode 201: advance Address by 14 to 0x400627 and Line by 0 to 34
  Set column to 27
  Extended opcode 4: set Discriminator to 3
  Special opcode 75: advance Address by 5 to 0x40062c and Line by 0 to 34
  Set column to 3
  Extended opcode 4: set Discriminator to 0
  Advance PC by constant 17 to 0x40063d
  Special opcode 63: advance Address by 4 to 0x400641 and Line by 2 to 36
  Advance PC by 2 to 0x400643
  Extended opcode 1: End of Sequence

Thus gdb gets confused - the pc is in the middle of line 34, but gdb doesn't
think so.

patch to follow.

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