From: "Christian Biesinger (Code Review)" <gerrit@gnutoolchain-gerrit.osci.io>
To: Christian Biesinger <cbiesinger@google.com>,
Tom Tromey <tromey@sourceware.org>,
gdb-patches@sourceware.org
Subject: [review v3] Precompute hash value for symbol_set_names
Date: Thu, 31 Oct 2019 00:23:00 -0000 [thread overview]
Message-ID: <20191031002337.2E19420AF6@gnutoolchain-gerrit.osci.io> (raw)
In-Reply-To: <gerrit.1572031795000.I044449e7eb60cffc1c43efd3412f2b485bd9faac@gnutoolchain-gerrit.osci.io>
Change URL: https://gnutoolchain-gerrit.osci.io/r/c/binutils-gdb/+/307
......................................................................
Precompute hash value for symbol_set_names
We can also compute the hash for the mangled name on a background
thread so make this function even faster (about a 7% speedup).
gdb/ChangeLog:
2019-10-03 Christian Biesinger <cbiesinger@google.com>
* minsyms.c (minimal_symbol_reader::install): Also compute the hash
of the mangled name on the background thread.
* symtab.c (symbol_set_names): Allow passing in the hash of the
linkage_name.
* symtab.h (symbol_set_names): Likewise.
Change-Id: I044449e7eb60cffc1c43efd3412f2b485bd9faac
---
M gdb/minsyms.c
M gdb/symtab.c
M gdb/symtab.h
3 files changed, 30 insertions(+), 6 deletions(-)
diff --git a/gdb/minsyms.c b/gdb/minsyms.c
index f9d1172..d2d8bf5 100644
--- a/gdb/minsyms.c
+++ b/gdb/minsyms.c
@@ -1258,6 +1258,12 @@
}
}
+struct computed_hash_values
+{
+ size_t name_length;
+ hashval_t mangled_name_hash;
+};
+
/* Build (or rebuild) the minimal symbol hash tables. This is necessary
after compacting or sorting the table since the entries move around
thus causing the internal minimal_symbol pointers to become jumbled. */
@@ -1370,6 +1376,8 @@
std::mutex demangled_mutex;
#endif
+ std::vector<computed_hash_values> hash_values (mcount);
+
msymbols = m_objfile->per_bfd->msymbols.get ();
gdb::parallel_for_each
(&msymbols[0], &msymbols[mcount],
@@ -1377,6 +1385,8 @@
{
for (minimal_symbol *msym = start; msym < end; ++msym)
{
+ size_t idx = msym - msymbols;
+ hash_values[idx].name_length = strlen (msym->name);
if (!msym->name_set)
{
/* This will be freed later, by symbol_set_names. */
@@ -1386,6 +1396,9 @@
(msym, demangled_name,
&m_objfile->per_bfd->storage_obstack);
msym->name_set = 1;
+
+ hash_values[idx].mangled_name_hash
+ = fast_hash (msym->name, hash_values[idx].name_length);
}
}
{
@@ -1396,8 +1409,14 @@
#endif
for (minimal_symbol *msym = start; msym < end; ++msym)
{
- symbol_set_names (msym, msym->name, false,
- m_objfile->per_bfd);
+ size_t idx = msym - msymbols;
+ symbol_set_names
+ (msym,
+ gdb::string_view(msym->name,
+ hash_values[idx].name_length),
+ false,
+ m_objfile->per_bfd,
+ hash_values[idx].mangled_name_hash);
}
}
});
diff --git a/gdb/symtab.c b/gdb/symtab.c
index 3502827..f5a759d 100644
--- a/gdb/symtab.c
+++ b/gdb/symtab.c
@@ -825,7 +825,8 @@
void
symbol_set_names (struct general_symbol_info *gsymbol,
gdb::string_view linkage_name, bool copy_name,
- struct objfile_per_bfd_storage *per_bfd)
+ struct objfile_per_bfd_storage *per_bfd,
+ gdb::optional<hashval_t> hash)
{
struct demangled_name_entry **slot;
@@ -853,9 +854,11 @@
create_demangled_names_hash (per_bfd);
struct demangled_name_entry entry (linkage_name);
+ if (!hash.has_value ())
+ hash = hash_demangled_name_entry (&entry);
slot = ((struct demangled_name_entry **)
- htab_find_slot (per_bfd->demangled_names_hash.get (),
- &entry, INSERT));
+ htab_find_slot_with_hash (per_bfd->demangled_names_hash.get (),
+ &entry, *hash, INSERT));
/* If this name is not in the hash table, add it. */
if (*slot == NULL
diff --git a/gdb/symtab.h b/gdb/symtab.h
index 4172a25..fe74204 100644
--- a/gdb/symtab.h
+++ b/gdb/symtab.h
@@ -522,7 +522,9 @@
(objfile)->per_bfd)
extern void symbol_set_names (struct general_symbol_info *symbol,
gdb::string_view linkage_name, bool copy_name,
- struct objfile_per_bfd_storage *per_bfd);
+ struct objfile_per_bfd_storage *per_bfd,
+ gdb::optional<hashval_t> hash
+ = gdb::optional<hashval_t> ());
/* Now come lots of name accessor macros. Short version as to when to
use which: Use SYMBOL_NATURAL_NAME to refer to the name of the
--
Gerrit-Project: binutils-gdb
Gerrit-Branch: master
Gerrit-Change-Id: I044449e7eb60cffc1c43efd3412f2b485bd9faac
Gerrit-Change-Number: 307
Gerrit-PatchSet: 3
Gerrit-Owner: Christian Biesinger <cbiesinger@google.com>
Gerrit-Reviewer: Tom Tromey <tromey@sourceware.org>
Gerrit-MessageType: newpatchset
next prev parent reply other threads:[~2019-10-31 0:23 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-10-25 19:30 [review] " Christian Biesinger (Code Review)
2019-10-29 19:28 ` Tom Tromey (Code Review)
2019-10-29 21:56 ` [review v2] " Christian Biesinger (Code Review)
2019-10-29 21:56 ` Christian Biesinger (Code Review)
2019-10-31 0:23 ` Christian Biesinger (Code Review) [this message]
2019-11-26 22:08 ` [review v4] " Tom Tromey (Code Review)
2019-11-26 22:23 ` [review v5] " Christian Biesinger (Code Review)
2019-11-26 22:24 ` Christian Biesinger (Code Review)
2019-11-27 18:03 ` Tom Tromey (Code Review)
2019-11-27 21:40 ` [pushed] " Sourceware to Gerrit sync (Code Review)
2019-11-27 21:40 ` Sourceware to Gerrit sync (Code Review)
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20191031002337.2E19420AF6@gnutoolchain-gerrit.osci.io \
--to=gerrit@gnutoolchain-gerrit.osci.io \
--cc=cbiesinger@google.com \
--cc=gdb-patches@sourceware.org \
--cc=tromey@sourceware.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).