public inbox for gdb-patches@sourceware.org
 help / color / mirror / Atom feed
* [PATCH] Fix ASAN failure in DWO code
@ 2024-01-09 19:51 Tom Tromey
  2024-01-09 20:29 ` Simon Marchi
  0 siblings, 1 reply; 2+ messages in thread
From: Tom Tromey @ 2024-01-09 19:51 UTC (permalink / raw)
  To: gdb-patches; +Cc: Tom Tromey

Simon pointed out that my recent change to the DWO code caused a
failure in ASAN testing.

The bug here was I updated the code to use a different search type in
the hash table; but then did not change the search code to use
htab_find_slot_with_hash.

Note that this bug would not be possible with my type-safe hash table
series, hint, hint.
---
 gdb/dwarf2/read.c | 31 +++++++++++++++++++++++--------
 1 file changed, 23 insertions(+), 8 deletions(-)

diff --git a/gdb/dwarf2/read.c b/gdb/dwarf2/read.c
index e055371dc6a..ca4dda5aea1 100644
--- a/gdb/dwarf2/read.c
+++ b/gdb/dwarf2/read.c
@@ -7783,16 +7783,24 @@ read_type_unit_scope (struct die_info *die, struct dwarf2_cu *cu)
    and DWP files (a file with the DWOs packaged up into one file), we treat
    DWP files as having a collection of virtual DWO files.  */
 
+/* A helper function to hash two file names.  This is a separate
+   function because the hash table uses a search with a different
+   type.  The second file may be NULL.  */
+
+static hashval_t
+hash_two_files (const char *one, const char *two)
+{
+  hashval_t hash = htab_hash_string (one);
+  if (two != nullptr)
+    hash += htab_hash_string (two);
+  return hash;
+}
+
 static hashval_t
 hash_dwo_file (const void *item)
 {
   const struct dwo_file *dwo_file = (const struct dwo_file *) item;
-  hashval_t hash;
-
-  hash = htab_hash_string (dwo_file->dwo_name.c_str ());
-  if (dwo_file->comp_dir != NULL)
-    hash += htab_hash_string (dwo_file->comp_dir);
-  return hash;
+  return hash_two_files (dwo_file->dwo_name.c_str (), dwo_file->comp_dir);
 }
 
 /* This is used when looking up entries in the DWO hash table.  */
@@ -7803,6 +7811,12 @@ struct dwo_file_search
   const char *dwo_name;
   /* Compilation directory to look for.  */
   const char *comp_dir;
+
+  /* Return a hash value compatible with the table.  */
+  hashval_t hash () const
+  {
+    return hash_two_files (dwo_name, comp_dir);
+  }
 };
 
 static int
@@ -7846,8 +7860,9 @@ lookup_dwo_file_slot (dwarf2_per_objfile *per_objfile,
 
   find_entry.dwo_name = dwo_name;
   find_entry.comp_dir = comp_dir;
-  slot = htab_find_slot (per_objfile->per_bfd->dwo_files.get (), &find_entry,
-			 INSERT);
+  slot = htab_find_slot_with_hash (per_objfile->per_bfd->dwo_files.get (),
+				   &find_entry, find_entry.hash (),
+				   INSERT);
 
   return slot;
 }
-- 
2.43.0


^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [PATCH] Fix ASAN failure in DWO code
  2024-01-09 19:51 [PATCH] Fix ASAN failure in DWO code Tom Tromey
@ 2024-01-09 20:29 ` Simon Marchi
  0 siblings, 0 replies; 2+ messages in thread
From: Simon Marchi @ 2024-01-09 20:29 UTC (permalink / raw)
  To: Tom Tromey, gdb-patches

On 1/9/24 14:51, Tom Tromey wrote:
> Simon pointed out that my recent change to the DWO code caused a
> failure in ASAN testing.
> 
> The bug here was I updated the code to use a different search type in
> the hash table; but then did not change the search code to use
> htab_find_slot_with_hash.
> 
> Note that this bug would not be possible with my type-safe hash table
> series, hint, hint.

I tested it locally, it seems fine to me.

Approved-By: Simon Marchi <simon.marchi@efficios.com>

Simon

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2024-01-09 20:29 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2024-01-09 19:51 [PATCH] Fix ASAN failure in DWO code Tom Tromey
2024-01-09 20:29 ` Simon Marchi

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).