From: Tom Tromey <tom@tromey.com>
To: gdb-patches@sourceware.org
Cc: Eli Zaretskii <eliz@gnu.org>
Subject: [PATCH v4 10/19] Add "maint set dwarf synchronous"
Date: Sun, 10 Dec 2023 14:41:18 -0700 [thread overview]
Message-ID: <20231210-t-bg-dwarf-reading-v4-10-b978c32fd12f@tromey.com> (raw)
In-Reply-To: <20231210-t-bg-dwarf-reading-v4-0-b978c32fd12f@tromey.com>
For testing, it's sometimes convenient to be able to request that
DWARF reading be done synchronously. This patch adds a new "maint"
setting for this purpose.
Reviewed-By: Eli Zaretskii <eliz@gnu.org>
---
gdb/NEWS | 4 ++++
gdb/doc/gdb.texinfo | 18 ++++++++++++++++++
gdb/dwarf2/read.c | 27 +++++++++++++++++++++++++++
3 files changed, 49 insertions(+)
diff --git a/gdb/NEWS b/gdb/NEWS
index c8166f08218..810fe617f52 100644
--- a/gdb/NEWS
+++ b/gdb/NEWS
@@ -9,6 +9,10 @@
* GDB index now contains information about the main function. This speeds up
startup when it is being used for some large binaries.
+* On hosts where threading is available, DWARF reading is now done in
+ the background, resulting in faster startup. This can be controlled
+ using "maint set dwarf synchronous".
+
* Changed commands
disassemble
diff --git a/gdb/doc/gdb.texinfo b/gdb/doc/gdb.texinfo
index 58685a77fd4..139484460c8 100644
--- a/gdb/doc/gdb.texinfo
+++ b/gdb/doc/gdb.texinfo
@@ -41764,6 +41764,24 @@ compilation units will be stored in memory longer, and more total
memory will be used. Setting it to zero disables caching, which will
slow down @value{GDBN} startup, but reduce memory consumption.
+@kindex maint set dwarf synchronous
+@kindex maint show dwarf synchronous
+@item maint set dwarf synchronous
+@itemx maint show dwarf synchronous
+Control whether DWARF is read asynchronously.
+
+On hosts where threading is available, the DWARF reader is mostly
+asynchronous with respect to the rest of @value{GDBN}. That is, the
+bulk of the reading is done in the background, and @value{GDBN} will
+only pause for completion of this task when absolutely necessary.
+
+When this setting is enabled, @value{GDBN} will instead wait for DWARF
+processing to complete before continuing.
+
+On hosts without threading, or where worker threads have been disabled
+at runtime, this setting has no effect, as DWARF reading is always
+done on the main thread, and is therefore always synchronous.
+
@kindex maint set dwarf unwinders
@kindex maint show dwarf unwinders
@item maint set dwarf unwinders
diff --git a/gdb/dwarf2/read.c b/gdb/dwarf2/read.c
index afadb05d9d0..b3f1c6b5dae 100644
--- a/gdb/dwarf2/read.c
+++ b/gdb/dwarf2/read.c
@@ -736,6 +736,18 @@ show_dwarf_max_cache_age (struct ui_file *file, int from_tty,
"DWARF compilation units is %s.\n"),
value);
}
+
+/* When true, wait for DWARF reading to be complete. */
+static bool dwarf_synchronous = false;
+
+/* "Show" callback for "maint set dwarf synchronous". */
+static void
+show_dwarf_synchronous (struct ui_file *file, int from_tty,
+ struct cmd_list_element *c, const char *value)
+{
+ gdb_printf (file, _("Whether DWARF reading is synchronous is %s.\n"),
+ value);
+}
\f
/* local function prototypes */
@@ -21832,6 +21844,21 @@ caching, which can slow down startup."),
&set_dwarf_cmdlist,
&show_dwarf_cmdlist);
+ add_setshow_boolean_cmd ("synchronous", class_obscure,
+ &dwarf_synchronous, _("\
+Set whether DWARF is read synchronously."), _("\
+Show whether DWARF is read synchronously."), _("\
+By default, DWARF information is read in worker threads,\n\
+and gdb will not generally wait for the reading to complete\n\
+before continuing with other work, for example presenting a\n\
+prompt to the user.\n\
+Enabling this setting will cause the DWARF reader to always wait\n\
+for debug info processing to be finished before gdb can proceed."),
+ nullptr,
+ show_dwarf_synchronous,
+ &set_dwarf_cmdlist,
+ &show_dwarf_cmdlist);
+
add_setshow_zuinteger_cmd ("dwarf-read", no_class, &dwarf_read_debug, _("\
Set debugging of the DWARF reader."), _("\
Show debugging of the DWARF reader."), _("\
--
2.43.0
next prev parent reply other threads:[~2023-12-10 21:41 UTC|newest]
Thread overview: 25+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-12-10 21:41 [PATCH v4 00/19] Index DWARF in the background Tom Tromey
2023-12-10 21:41 ` [PATCH v4 01/19] Don't use objfile::intern in DWO code Tom Tromey
2024-01-09 18:44 ` Simon Marchi
2024-01-09 19:12 ` Tom Tromey
2024-01-09 19:51 ` Tom Tromey
2023-12-10 21:41 ` [PATCH v4 02/19] Pre-read DWZ section data Tom Tromey
2023-12-10 21:41 ` [PATCH v4 03/19] Add a couple of bfd_cache_close calls Tom Tromey
2023-12-10 21:41 ` [PATCH v4 04/19] Add thread-safety to gdb's BFD wrappers Tom Tromey
2023-12-10 21:41 ` [PATCH v4 05/19] Refactor complaint thread-safety approach Tom Tromey
2023-12-10 21:41 ` [PATCH v4 06/19] Add deferred_warnings parameter to read_addrmap_from_aranges Tom Tromey
2023-12-10 21:41 ` [PATCH v4 07/19] Add quick_symbol_functions::compute_main_name Tom Tromey
2023-12-10 21:41 ` [PATCH v4 08/19] Add gdb::task_group Tom Tromey
2023-12-18 14:55 ` Tom Tromey
2023-12-10 21:41 ` [PATCH v4 09/19] Move cooked_index_storage to cooked-index.h Tom Tromey
2023-12-10 21:41 ` Tom Tromey [this message]
2023-12-10 21:41 ` [PATCH v4 11/19] Change how cooked index waits for threads Tom Tromey
2023-12-10 21:41 ` [PATCH v4 12/19] Do more DWARF reading in the background Tom Tromey
2023-12-10 21:41 ` [PATCH v4 13/19] Simplify the public DWARF API Tom Tromey
2023-12-10 21:41 ` [PATCH v4 14/19] Remove two quick_symbol_functions methods Tom Tromey
2023-12-10 21:41 ` [PATCH v4 15/19] Change current_language to be a macro Tom Tromey
2023-12-10 21:41 ` [PATCH v4 16/19] Lazy language setting Tom Tromey
2023-12-10 21:41 ` [PATCH v4 17/19] Optimize lookup_minimal_symbol_text Tom Tromey
2023-12-10 21:41 ` [PATCH v4 18/19] Avoid language-based lookups in startup path Tom Tromey
2023-12-10 21:41 ` [PATCH v4 19/19] Back out some parallel_for_each features Tom Tromey
2024-01-09 1:08 ` [PATCH v4 00/19] Index DWARF in the background Tom Tromey
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=20231210-t-bg-dwarf-reading-v4-10-b978c32fd12f@tromey.com \
--to=tom@tromey.com \
--cc=eliz@gnu.org \
--cc=gdb-patches@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).