[RFC] Use partial symbol table to find language for main
Commit Message
Hi,
When language is set to auto, part of loading an executable is to update the
language accordingly. This is implemented by set_initial_language.
The implementation of set_initial_language works as follows:
- check if any objfile in the progspace has name_of_main/language_of_main
set, and if so, use the first one found. [ This is what you get f.i. when
using dwarf with DW_AT_main_subprogram. ]
- otherwise, check for known names in the minimal symbols, and either:
- use the associated language if any (f.i. for ada), or
- lookup the symbol in the symtab for the name and use the symbol language
(f.i. for c/c++).
The symbol lookup can be slow though.
In the case of the cc1 binary from PR23710 comment 1, getting to the initial
prompt takes ~8s:
...
$ time.sh gdb cc1 -batch -ex "show language"
The current source language is "auto; currently c++".
maxmem: 1272260
real: 8.05
user: 7.73
system: 0.38
...
but if we skip guessing the initial language by setting it instead, it takes
only ~4s:
...
$ time.sh gdb -iex "set language c++" cc1 -batch -ex "show language"
The current source language is "c++".
maxmem: 498272
real: 3.99
user: 3.90
system: 0.15
...
In both cases, we load the partial symbols for the executable, but in the
first case only we also do a lookup of main, which causes the corresponding
partial symtab to be expanded into a full symtab.
Ideally, we'd like to get the language of the symbol without triggering
expansion into a full symtab, and get the speedup without having to set the
language manually.
There's a related fixme in the header comment of set_initial_language:
...
/* Set the initial language.
FIXME: A better solution would be to record the language in the
psymtab when reading partial symbols, and then use it (if known) to
set the language. This would be a win for formats that encode the
language in an easily discoverable place, such as DWARF. For
stabs, we can jump through hoops looking for specially named
symbols or try to intuit the language from the specific type of
stabs we find, but we can't do that until later when we read in
full symbols. */
void
set_initial_language (void)
...
Since we're already tracking the language of partial symbols, use this to set
the language for the main symbol.
Note that this search in partial symbol tables is not guaranteed to yield the
same result as the lookup_symbol_in_language call currently done in
set_initial_language. I'm not sure if that matters.
Build and reg-tested on x86_64-linux.
This requires an update to test-case gdb.base/break-interp.exp to work around
PR25613 - "[gdb] verbose message appears not at start of line during
backtrace".
Any comments?
Thanks,
- Tom
[gdb] Use partial symbol table to find language for main
gdb/ChangeLog:
2020-03-01 Tom de Vries <tdevries@suse.de>
* psymtab.c (psym_lookup_symbol_2, find_psym_language): New function.
* psymtab.h (find_psym_language): Declare.
* symtab.c (find_main_name): Use find_psym_language.
* symfile.c (set_initial_language): Remove fixme comment.
gdb/testsuite/ChangeLog:
2020-03-01 Tom de Vries <tdevries@suse.de>
* gdb.base/break-interp.exp: Set verbose back to off.
* gdb.base/c-linkage-name.exp (verify_psymtab_expanded): Move ...
* lib/gdb.exp: ... here.
* gdb.base/main-psymtab.exp: New file.
---
gdb/psymtab.c | 47 +++++++++++++++++++++++++++++++
gdb/psymtab.h | 4 +++
gdb/symfile.c | 11 +-------
gdb/symtab.c | 9 ++++++
gdb/testsuite/gdb.base/break-interp.exp | 2 ++
gdb/testsuite/gdb.base/c-linkage-name.exp | 13 ---------
gdb/testsuite/gdb.base/main-psymtab.exp | 34 ++++++++++++++++++++++
gdb/testsuite/lib/gdb.exp | 13 +++++++++
8 files changed, 110 insertions(+), 23 deletions(-)
Comments
>>>>> "Tom" == Tom de Vries <tdevries@suse.de> writes:
Tom> Ideally, we'd like to get the language of the symbol without triggering
Tom> expansion into a full symtab, and get the speedup without having to set the
Tom> language manually.
This seems like a good idea.
Tom> +extern enum language
Tom> +find_psym_language (block_enum block_index, const char *name,
Tom> + const domain_enum domain);
FYI we don't normally line break after the type for declarations, only
definitions.
Tom> /* The languages above didn't identify the name of the main procedure.
Tom> Fallback to "main". */
Tom> +
Tom> + /* Try to find language for main in psymtabs. */
Tom> + enum language lang = find_psym_language (GLOBAL_BLOCK, "main", VAR_DOMAIN);
Generic code shouldn't call directly into the psymtab code. While most
debug readers create psymtabs, they aren't required to. For example,
the DWARF index code bypasses psymtabs entirely -- but this call might
cause the DWARF reader to create psymtabs in this case.
Instead the way to do this is add a new function pointer to
struct quick_symbol_functions, then implement it in the various debug
readers.
Tom
@@ -517,6 +517,53 @@ psym_lookup_symbol (struct objfile *objfile,
return stab_best;
}
+/* Variant of psym_lookup_symbol that return a partial symbol. */
+
+static struct partial_symbol *
+psym_lookup_symbol_2 (struct objfile *objfile,
+ block_enum block_index, const char *name,
+ const domain_enum domain)
+{
+ if (objfile->sf == NULL)
+ return NULL;
+ const int psymtab_index = (block_index == GLOBAL_BLOCK ? 1 : 0);
+
+ lookup_name_info lookup_name (name, symbol_name_match_type::FULL);
+
+ for (partial_symtab *ps : require_partial_symbols (objfile, true))
+ {
+ struct partial_symbol *psym;
+ if (ps->readin_p ())
+ continue;
+
+ psym = lookup_partial_symbol (objfile, ps, name, psymtab_index, domain);
+ if (psym)
+ return psym;
+ }
+
+ return NULL;
+}
+
+/* Find the language for partial symbol with NAME. */
+
+enum language
+find_psym_language (block_enum block_index, const char *name,
+ const domain_enum domain)
+{
+ for (objfile *objfile : current_program_space->objfiles ())
+ {
+ struct partial_symbol *psym
+ = psym_lookup_symbol_2 (objfile, block_index, name, domain);
+ if (!psym)
+ continue;
+ enum language lang = psym->ginfo.language ();
+ if (lang != language_unknown)
+ return lang;
+ }
+
+ return language_unknown;
+}
+
/* Returns true if PSYM matches LOOKUP_NAME. */
static bool
@@ -151,4 +151,8 @@ extern const struct quick_symbol_functions dwarf2_debug_names_functions;
extern psymtab_storage::partial_symtab_range require_partial_symbols
(struct objfile *objfile, bool verbose);
+extern enum language
+find_psym_language (block_enum block_index, const char *name,
+ const domain_enum domain);
+
#endif /* PSYMTAB_H */
@@ -1667,16 +1667,7 @@ symbol_file_command (const char *args, int from_tty)
}
}
-/* Set the initial language.
-
- FIXME: A better solution would be to record the language in the
- psymtab when reading partial symbols, and then use it (if known) to
- set the language. This would be a win for formats that encode the
- language in an easily discoverable place, such as DWARF. For
- stabs, we can jump through hoops looking for specially named
- symbols or try to intuit the language from the specific type of
- stabs we find, but we can't do that until later when we read in
- full symbols. */
+/* Set the initial language. */
void
set_initial_language (void)
@@ -6090,6 +6090,15 @@ find_main_name (void)
/* The languages above didn't identify the name of the main procedure.
Fallback to "main". */
+
+ /* Try to find language for main in psymtabs. */
+ enum language lang = find_psym_language (GLOBAL_BLOCK, "main", VAR_DOMAIN);
+ if (lang != language_unknown)
+ {
+ set_main_name ("main", lang);
+ return;
+ }
+
set_main_name ("main", language_unknown);
}
@@ -251,6 +251,7 @@ proc test_core {file displacement} {
if ![regexp {^(NONE|FOUND-.*)$} $displacement] {
fail $test_displacement
}
+ gdb_test_no_output "set verbose off"
gdb_test "bt" "#\[0-9\]+ +\[^\r\n\]*\\mlibfunc\\M\[^\r\n\]*\r\n#\[0-9\]+ +\[^\r\n\]*\\mmain\\M.*" "core main bt"
}
@@ -301,6 +302,7 @@ proc test_attach_gdb {file pid displacement prefix} {
if ![regexp {^(NONE|FOUND-.*)$} $displacement] {
fail $test_displacement
}
+ gdb_test_no_output "set verbose off"
gdb_test "bt" "#\[0-9\]+ +\[^\r\n\]*\\mlibfunc\\M\[^\r\n\]*\r\n#\[0-9\]+ +\[^\r\n\]*\\mmain\\M.*" "attach main bt"
gdb_exit
@@ -27,19 +27,6 @@ if { [gdb_compile "${sources}" "${binfile}" executable {debug}] != "" } {
clean_restart ${binfile}
-# Verify that partial symtab expansion for $filename has state $readin
-
-proc verify_psymtab_expanded { filename readin } {
- set cmd "maint info psymtab"
- set test "$cmd: $filename: $readin"
- set re [multi_line \
- " \{ psymtab \[^\r\n\]*$filename\[^\r\n\]*" \
- " readin $readin" \
- ".*"]
-
- gdb_test $cmd $re $test
-}
-
# Verify that partial symtab expansion has not taken place for
# c-linkage-name-2.c.
new file mode 100644
@@ -0,0 +1,34 @@
+# Copyright 2020 Free Software Foundation, Inc.
+
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program. If not, see <http://www.gnu.org/licenses/>.
+
+standard_testfile persistent-lang.cc
+
+if {[build_executable "failed to prepare" $testfile $srcfile debug]} {
+ return -1
+}
+
+clean_restart
+
+set auto_cpp \
+ {The current source language is "auto; currently c\+\+"\.}
+
+gdb_load ${binfile}
+gdb_test "show language" $auto_cpp \
+ "language auto/c++ after load"
+
+# Verify that partial symtab expansion has not taken place for
+# persistent-lang.cc
+
+verify_psymtab_expanded persistent-lang.cc no
@@ -6862,5 +6862,18 @@ gdb_caching_proc skip_ctf_tests {
} executable "additional_flags=-gt"]
}
+# Verify that partial symtab expansion for $filename has state $readin.
+
+proc verify_psymtab_expanded { filename readin } {
+ set cmd "maint info psymtab"
+ set test "$cmd: $filename: $readin"
+ set re [multi_line \
+ " \{ psymtab \[^\r\n\]*$filename\[^\r\n\]*" \
+ " readin $readin" \
+ ".*"]
+
+ gdb_test $cmd $re $test
+}
+
# Always load compatibility stuff.
load_lib future.exp