diff mbox series

[3/5,gdb/symtab] Revert "Change handling of DW_TAG_enumeration_type in DWARF scanner"

Message ID 20240912115242.17062-3-tdevries@suse.de
State Committed
Headers
Series [1/5,gdb/testsuite] Add gdb.dwarf2/enum-type-c++.exp, regression test for PR31900. |

Checks

Context Check Description
linaro-tcwg-bot/tcwg_gdb_build--master-aarch64 success Build passed
linaro-tcwg-bot/tcwg_gdb_build--master-arm success Build passed
linaro-tcwg-bot/tcwg_gdb_check--master-aarch64 success Test passed

Commit Message

Tom de Vries Sept. 12, 2024, 11:52 a.m. UTC 
  After adding dwarf assembly to test-case gdb.dwarf2/enum-type.exp that adds
this debug info:
...
 <1><11f>: Abbrev Number: 3 (DW_TAG_enumeration_type)
    <120>   DW_AT_specification: <0x130>
 <2><124>: Abbrev Number: 4 (DW_TAG_enumerator)
    <125>   DW_AT_name        : val1
    <12a>   DW_AT_const_value : 1
 <2><12b>: Abbrev Number: 0
 <1><12c>: Abbrev Number: 5 (DW_TAG_namespace)
    <12d>   DW_AT_name        : ns
 <2><130>: Abbrev Number: 6 (DW_TAG_enumeration_type)
    <131>   DW_AT_name        : e
    <133>   DW_AT_type        : <0x118>
    <137>   DW_AT_declaration : 1
...
I run into an assertion failure:
...
(gdb) file enum-type^M
Reading symbols from enum-type...^M
cooked-index.h:214: internal-error: get_parent: \
  Assertion `(flags & IS_PARENT_DEFERRED) == 0' failed.^M
...

This was reported in PR32160 comment 1.

This is a regression since commit 4e417d7bb1c ("Change handling of
DW_TAG_enumeration_type in DWARF scanner").

Fix this by reverting the commit.

[ Also drop the kfails for PR31900 and PR32158, which are regressions by that
same commit. ]

That allows us to look at the output of "maint print objfiles", and for val1
we get an entry without parent:
...
    [27] ((cooked_index_entry *) 0x7fbbb4002ef0)
    name:       val1
    canonical:  val1
    qualified:  val1
    DWARF tag:  DW_TAG_enumerator
    flags:      0x0 []
    DIE offset: 0x124
    parent:     ((cooked_index_entry *) 0)
...
which is incorrect, as noted in that same comment, but an improvement over the
assertion failure, and I don't think that ever worked.  This is to be
addressed in a follow-up patch.

Reverting the commit begs the question: what was it trying to fix in the first
place, and do we need a different fix?  I've investigated this and filed
PR32160 to track this.

My guess is that the commit was based on a misunderstand of what we track
in cooked_indexer::m_die_range_map.

Each DIE has two types of parent DIEs:
- a DIE that is the parent as indicated by the tree structure in which DIEs
  occur, and
- a DIE that represent the parent scope.

In most cases, these two are the same, but some times they're not.

The debug info above demonstrates such a case.  The DIE at 0x11f:
- has a tree-parent: the DIE representing the CU, and
- has a scope-parent: DIE 0x12c representing namespace ns.

In cooked_indexer::m_die_range_map, we track scope-parents, and the commit
tried to add a tree-parent instead.

So, I don't think we need a different fix, and propose we backport the reversal
for gdb 15.2.

Tested on x86_64-linux.

Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=31900
Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=32158
Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=32160
---
 gdb/dwarf2/read.c                          | 16 +++++-----
 gdb/testsuite/gdb.dwarf2/enum-type-c++.exp |  2 --
 gdb/testsuite/gdb.dwarf2/enum-type.exp     | 35 ++++++++++++++++++++++
 3 files changed, 44 insertions(+), 9 deletions(-)

Comments

Tom Tromey Sept. 13, 2024, 7:11 p.m. UTC | #1 
>>>>> "Tom" == Tom de Vries <tdevries@suse.de> writes:

Tom> Reverting the commit begs the question: what was it trying to fix in the first
Tom> place, and do we need a different fix?  I've investigated this and filed
Tom> PR32160 to track this.

Tom> My guess is that the commit was based on a misunderstand of what we track
Tom> in cooked_indexer::m_die_range_map.

It's been a while but my recollection is that, if you examine the DIE
range map while the cooked index is being built, you'll see entries in
it that are wrong -- they attribute the wrong context offset to a DIE
range.

I don't really recall any more than that.

Tom> +	    DW_TAG_enumeration_type {
Tom> +		{DW_AT_specification :$forward}
Tom> +	    } {

Ugh DWARF.

Anyway I didn't really understand why the backout patch adds a new test,
but I suppose it's not important.

Approved-By: Tom Tromey <tom@tromey.com>

Tom
Tom de Vries Sept. 14, 2024, 11:51 a.m. UTC | #2 
On 9/13/24 21:11, Tom Tromey wrote:
>>>>>> "Tom" == Tom de Vries <tdevries@suse.de> writes:
> 
> Tom> Reverting the commit begs the question: what was it trying to fix in the first
> Tom> place, and do we need a different fix?  I've investigated this and filed
> Tom> PR32160 to track this.
> 
> Tom> My guess is that the commit was based on a misunderstand of what we track
> Tom> in cooked_indexer::m_die_range_map.
> 
> It's been a while but my recollection is that, if you examine the DIE
> range map while the cooked index is being built, you'll see entries in
> it that are wrong -- they attribute the wrong context offset to a DIE
> range.
> 

Hi Tom,

thanks for the review(s).

> I don't really recall any more than that.
> 
> Tom> +	    DW_TAG_enumeration_type {
> Tom> +		{DW_AT_specification :$forward}
> Tom> +	    } {
> 
> Ugh DWARF.
> 
> Anyway I didn't really understand why the backout patch adds a new test,
> but I suppose it's not important.
> 

Well, I could have added it as a separate patch before, but adding a 
kfail for an assertion failure is a bit more complicated.

I could also have added it as a separate patch afterwards.

I didn't see the problem with adding it to the reversal patch, so I 
guess I must be missing something.

Anyway, I'll push the 3 approved patches shortly.

Thanks,
- Tom

> Approved-By: Tom Tromey <tom@tromey.com>
> 
> Tom
diff mbox series

Patch

diff --git a/gdb/dwarf2/read.c b/gdb/dwarf2/read.c
index e0dcf940402..bbd6bfbaf45 100644
--- a/gdb/dwarf2/read.c
+++ b/gdb/dwarf2/read.c
@@ -16471,12 +16471,6 @@  cooked_indexer::index_dies (cutu_reader *reader,
 	  flags &= ~IS_STATIC;
 	  flags |= parent_entry->flags & IS_STATIC;
 	}
-      /* If the parent is an enum, but not an enum class, then use the
-	 grandparent instead.  */
-      if (this_parent_entry != nullptr
-	  && this_parent_entry->tag == DW_TAG_enumeration_type
-	  && !is_enum_class)
-	this_parent_entry = this_parent_entry->get_parent ();
 
       if (abbrev->tag == DW_TAG_namespace
 	  && m_language == language_cplus
@@ -16536,7 +16530,15 @@  cooked_indexer::index_dies (cutu_reader *reader,
 	      break;
 
 	    case DW_TAG_enumeration_type:
-	      info_ptr = recurse (reader, info_ptr, this_entry, fully);
+	      /* We need to recurse even for an anonymous enumeration.
+		 Which scope we record as the parent scope depends on
+		 whether we're reading an "enum class".  If so, we use
+		 the enum itself as the parent, yielding names like
+		 "enum_class::enumerator"; otherwise we inject the
+		 names into our own parent scope.  */
+	      info_ptr = recurse (reader, info_ptr,
+				  is_enum_class ? this_entry : parent_entry,
+				  fully);
 	      continue;
 
 	    case DW_TAG_module:
diff --git a/gdb/testsuite/gdb.dwarf2/enum-type-c++.exp b/gdb/testsuite/gdb.dwarf2/enum-type-c++.exp
index a3dfb4d2f74..8bf737ec506 100644
--- a/gdb/testsuite/gdb.dwarf2/enum-type-c++.exp
+++ b/gdb/testsuite/gdb.dwarf2/enum-type-c++.exp
@@ -32,7 +32,6 @@  require {string equal [have_index $binfile] ""}
 set re_ws "\[ \t\]"
 
 # Regression test for PR31900.
-setup_kfail "gdb/31900" *-*-*
 set val1 ns::A::val1
 gdb_test_lines "maint print objfiles" \
     "val1 has a parent" \
@@ -44,7 +43,6 @@  gdb_test_lines "maint print objfiles" \
 gdb_test "print $val1" " = $val1"
 
 # Regression test for PR32158.
-setup_kfail "gdb/32158" *-*-*
 set val2 ns::ec::val2
 gdb_test_lines "maint print objfiles" \
     "val2 has correct parent" \
diff --git a/gdb/testsuite/gdb.dwarf2/enum-type.exp b/gdb/testsuite/gdb.dwarf2/enum-type.exp
index 394d287738b..b2b3dc6c828 100644
--- a/gdb/testsuite/gdb.dwarf2/enum-type.exp
+++ b/gdb/testsuite/gdb.dwarf2/enum-type.exp
@@ -65,6 +65,41 @@  Dwarf::assemble $asm_file {
 	    }
 	}
     }
+
+    cu {} {
+	DW_TAG_compile_unit {
+		{DW_AT_language @DW_LANG_C_plus_plus}
+		{DW_AT_name	tmp.c}
+		{DW_AT_comp_dir /tmp}
+	} {
+	    declare_labels integer_label forward
+
+	    integer_label: DW_TAG_base_type {
+		{DW_AT_byte_size 4 DW_FORM_sdata}
+		{DW_AT_encoding	 @DW_ATE_signed}
+		{DW_AT_name	 int}
+	    }
+
+	    DW_TAG_enumeration_type {
+		{DW_AT_specification :$forward}
+	    } {
+		DW_TAG_enumerator {
+		    {DW_AT_name val1}
+		    {DW_AT_const_value 1 DW_FORM_sdata}
+		}
+	    }
+
+	    DW_TAG_namespace {
+		{DW_AT_name ns}
+	    } {
+		forward: DW_TAG_enumeration_type {
+		    {DW_AT_name e}
+		    {DW_AT_type :$integer_label}
+		    {DW_AT_declaration 1 flag}
+		}
+	    }
+	}
+    }
 }
 
 if { [prepare_for_testing "failed to prepare" ${testfile} \