diff options
author | Martin Galvan <martin.galvan@tallertechnologies.com> | 2016-12-08 18:59:02 +0530 |
---|---|---|
committer | Siddhesh Poyarekar <siddhesh@sourceware.org> | 2016-12-08 18:59:02 +0530 |
commit | 23b5cae1af04f2d912910fdaf73cb482265798c1 (patch) | |
tree | d7d463b80da7c33ada5073d1e09bc889639b82e7 /scripts | |
parent | 297635d82bf5ff55899f694a5261ffd97636df98 (diff) | |
download | glibc-23b5cae1af04f2d912910fdaf73cb482265798c1.tar.gz |
Add pretty printers for the NPTL lock types
This patch adds pretty printers for the following NPTL types:
- pthread_mutex_t
- pthread_mutexattr_t
- pthread_cond_t
- pthread_condattr_t
- pthread_rwlock_t
- pthread_rwlockattr_t
To load the pretty printers into your gdb session, do the following:
python
import sys
sys.path.insert(0, '/path/to/glibc/build/nptl/pretty-printers')
end
source /path/to/glibc/source/pretty-printers/nptl-printers.py
You can check which printers are registered and enabled by issuing the
'info pretty-printer' gdb command. Printers should trigger automatically when
trying to print a variable of one of the types mentioned above.
The printers are architecture-independent, and were tested on an AMD64 running
Ubuntu 14.04 and an x86 VM running Fedora 24.
In order to work, the printers need to know the values of various flags that
are scattered throughout pthread.h and pthreadP.h as enums and #defines. Since
replicating these constants in the printers file itself would create a
maintenance burden, I wrote a script called gen-py-const.awk that Makerules uses
to extract the constants. This script is pretty much the same as gen-as-const.awk,
except it doesn't cast the constant values to 'long' and is thorougly documented.
The constants need only to be enumerated in a .pysym file, which is then referenced
by a Make variable called gen-py-const-headers.
As for the install directory, I discussed this with Mike Frysinger and Siddhesh
Poyarekar, and we agreed that it can be handled in a separate patch, and shouldn't
block merging of this one.
In addition, I've written a series of test cases for the pretty printers.
Each lock type (mutex, condvar and rwlock) has two test programs, one for itself
and other for its related 'attributes' object. Each test program in turn has a
PExpect-based Python script that drives gdb and compares its output to the
expected printer's. The tests run on the glibc host, which is assumed to have
both gdb and PExpect; if either is absent the tests will fail with code 77
(UNSUPPORTED). For cross-testing you should use cross-test-ssh.sh as test-wrapper.
I've tested the printers on both native builds and a cross build using a Beaglebone
Black running Debian, with the build system's filesystem shared with the board
through NFS.
Finally, I've written a README that explains all this and more.
* INSTALL: Regenerated.
* Makeconfig: Add comments and whitespace to make the control flow
clearer.
(+link-printers-tests, +link-pie-printers-tests, CFLAGS-printers-tests,
installed-rtld-LDFLAGS, built-rtld-LDFLAGS, link-libc-rpath,
link-libc-tests-after-rpath-link, link-libc-printers-tests): New.
(rtld-LDFLAGS, rtld-tests-LDFLAGS, link-libc-tests-rpath-link,
link-libc-tests): Use the new variables as required.
* Makerules ($(py-const)): New rule.
generated: Add $(py-const).
* README.pretty-printers: New file.
* Rules (tests-printers-programs, tests-printers-out, py-env): New.
(others): Depend on $(py-const).
(tests): Depend on $(tests-printers-programs) or $(tests-printers-out),
as required. Pass $(tests-printers) to merge-test-results.sh.
* manual/install.texi: Add requirements for testing the pretty printers.
* nptl/Makefile (gen-py-const-headers, pretty-printers, tests-printers,
CFLAGS-test-mutexattr-printers.c CFLAGS-test-mutex-printers.c,
CFLAGS-test-condattr-printers.c, CFLAGS-test-cond-printers.c,
CFLAGS-test-rwlockattr-printers.c CFLAGS-test-rwlock-printers.c,
tests-printers-libs): Define.
* nptl/nptl-printers.py: New file.
* nptl/nptl_lock_constants.pysym: Likewise.
* nptl/test-cond-printers.c: Likewise.
* nptl/test-cond-printers.py: Likewise.
* nptl/test-condattr-printers.c: Likewise.
* nptl/test-condattr-printers.py: Likewise.
* nptl/test-mutex-printers.c: Likewise.
* nptl/test-mutex-printers.py: Likewise.
* nptl/test-mutexattr-printers.c: Likewise.
* nptl/test-mutexattr-printers.py: Likewise.
* nptl/test-rwlock-printers.c: Likewise.
* nptl/test-rwlock-printers.py: Likewise.
* nptl/test-rwlockattr-printers.c: Likewise.
* nptl/test-rwlockattr-printers.py: Likewise.
* scripts/gen-py-const.awk: Likewise.
* scripts/test_printers_common.py: Likewise.
* scripts/test_printers_exceptions.py: Likewise.
Diffstat (limited to 'scripts')
-rw-r--r-- | scripts/gen-py-const.awk | 118 | ||||
-rw-r--r-- | scripts/test_printers_common.py | 364 | ||||
-rw-r--r-- | scripts/test_printers_exceptions.py | 61 |
3 files changed, 543 insertions, 0 deletions
diff --git a/scripts/gen-py-const.awk b/scripts/gen-py-const.awk new file mode 100644 index 0000000000..4586f59d89 --- /dev/null +++ b/scripts/gen-py-const.awk @@ -0,0 +1,118 @@ +# Script to generate constants for Python pretty printers. +# +# Copyright (C) 2016 Free Software Foundation, Inc. +# This file is part of the GNU C Library. +# +# The GNU C Library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2.1 of the License, or (at your option) any later version. +# +# The GNU C Library 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 +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with the GNU C Library; if not, see +# <http://www.gnu.org/licenses/>. + +# This script is a smaller version of the clever gen-asm-const.awk hack used to +# generate ASM constants from .sym files. We'll use this to generate constants +# for Python pretty printers. +# +# The input to this script are .pysym files that look like: +# #C_Preprocessor_Directive... +# NAME1 +# NAME2 expression... +# +# A line giving just a name implies an expression consisting of just that name. +# Comments start with '--'. +# +# The output of this script is a 'dummy' function containing 'asm' declarations +# for each non-preprocessor line in the .pysym file. The expression values +# will appear as input operands to the 'asm' declaration. For example, if we +# have: +# +# /* header.h */ +# #define MACRO 42 +# +# struct S { +# char c1; +# char c2; +# char c3; +# }; +# +# enum E { +# ZERO, +# ONE +# }; +# +# /* symbols.pysym */ +# #include <stddef.h> +# #include "header.h" +# -- This is a comment +# MACRO +# C3_OFFSET offsetof(struct S, c3) +# E_ONE ONE +# +# the output will be: +# +# #include <stddef.h> +# #include "header.h" +# void dummy(void) +# { +# asm ("@name@MACRO@value@%0@" : : "i" (MACRO)); +# asm ("@name@C3_OFFSET@value@%0@" : : "i" (offsetof(struct S, c3))); +# asm ("@name@E_ONE@value@%0@" : : "i" (ONE)); +# } +# +# We'll later feed this output to gcc -S. Since '-S' tells gcc to compile but +# not assemble, gcc will output something like: +# +# dummy: +# ... +# @name@MACRO@value@$42@ +# @name@C3_OFFSET@value@$2@ +# @name@E_ONE@value@$1@ +# +# Finally, we can process that output to extract the constant values. +# Notice gcc may prepend a special character such as '$' to each value. + +# found_symbol indicates whether we found a non-comment, non-preprocessor line. +BEGIN { found_symbol = 0 } + +# C preprocessor directives go straight through. +/^#/ { print; next; } + +# Skip comments. +/--/ { next; } + +# Trim leading whitespace. +{ sub(/^[[:blank:]]*/, ""); } + +# If we found a non-comment, non-preprocessor line, print the 'dummy' function +# header. +NF > 0 && !found_symbol { + print "void dummy(void)\n{"; + found_symbol = 1; +} + +# If the line contains just a name, duplicate it so we can use that name +# as the value of the expression. +NF == 1 { sub(/^.*$/, "& &"); } + +# If a line contains a name and an expression... +NF > 1 { + name = $1; + + # Remove any characters before the second field. + sub(/^[^[:blank:]]+[[:blank:]]+/, ""); + + # '$0' ends up being everything that appeared after the first field + # separator. + printf " asm (\"@name@%s@value@%0@\" : : \"i\" (%s));\n", name, $0; +} + +# Close the 'dummy' function. +END { if (found_symbol) print "}"; } diff --git a/scripts/test_printers_common.py b/scripts/test_printers_common.py new file mode 100644 index 0000000000..c79d7e3be2 --- /dev/null +++ b/scripts/test_printers_common.py @@ -0,0 +1,364 @@ +# Common functions and variables for testing the Python pretty printers. +# +# Copyright (C) 2016 Free Software Foundation, Inc. +# This file is part of the GNU C Library. +# +# The GNU C Library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2.1 of the License, or (at your option) any later version. +# +# The GNU C Library 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 +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with the GNU C Library; if not, see +# <http://www.gnu.org/licenses/>. + +"""These tests require PExpect 4.0 or newer. + +Exported constants: + PASS, FAIL, UNSUPPORTED (int): Test exit codes, as per evaluate-test.sh. +""" + +import os +import re +from test_printers_exceptions import * + +PASS = 0 +FAIL = 1 +UNSUPPORTED = 77 + +gdb_bin = 'gdb' +gdb_options = '-q -nx' +gdb_invocation = '{0} {1}'.format(gdb_bin, gdb_options) +pexpect_min_version = 4 +gdb_min_version = (7, 8) +encoding = 'utf-8' + +try: + import pexpect +except ImportError: + print('PExpect 4.0 or newer must be installed to test the pretty printers.') + exit(UNSUPPORTED) + +pexpect_version = pexpect.__version__.split('.')[0] + +if int(pexpect_version) < pexpect_min_version: + print('PExpect 4.0 or newer must be installed to test the pretty printers.') + exit(UNSUPPORTED) + +if not pexpect.which(gdb_bin): + print('gdb 7.8 or newer must be installed to test the pretty printers.') + exit(UNSUPPORTED) + +timeout = 5 +TIMEOUTFACTOR = os.environ.get('TIMEOUTFACTOR') + +if TIMEOUTFACTOR: + timeout = int(TIMEOUTFACTOR) + +try: + # Check the gdb version. + version_cmd = '{0} --version'.format(gdb_invocation, timeout=timeout) + gdb_version_out = pexpect.run(version_cmd, encoding=encoding) + + # The gdb version string is "GNU gdb <PKGVERSION><version>", where + # PKGVERSION can be any text. We assume that there'll always be a space + # between PKGVERSION and the version number for the sake of the regexp. + version_match = re.search(r'GNU gdb .* ([1-9]+)\.([0-9]+)', gdb_version_out) + + if not version_match: + print('The gdb version string (gdb -v) is incorrectly formatted.') + exit(UNSUPPORTED) + + gdb_version = (int(version_match.group(1)), int(version_match.group(2))) + + if gdb_version < gdb_min_version: + print('gdb 7.8 or newer must be installed to test the pretty printers.') + exit(UNSUPPORTED) + + # Check if gdb supports Python. + gdb_python_cmd = '{0} -ex "python import os" -batch'.format(gdb_invocation, + timeout=timeout) + gdb_python_error = pexpect.run(gdb_python_cmd, encoding=encoding) + + if gdb_python_error: + print('gdb must have python support to test the pretty printers.') + exit(UNSUPPORTED) + + # If everything's ok, spawn the gdb process we'll use for testing. + gdb = pexpect.spawn(gdb_invocation, echo=False, timeout=timeout, + encoding=encoding) + gdb_prompt = u'\(gdb\)' + gdb.expect(gdb_prompt) + +except pexpect.ExceptionPexpect as exception: + print('Error: {0}'.format(exception)) + exit(FAIL) + +def test(command, pattern=None): + """Sends 'command' to gdb and expects the given 'pattern'. + + If 'pattern' is None, simply consumes everything up to and including + the gdb prompt. + + Args: + command (string): The command we'll send to gdb. + pattern (raw string): A pattern the gdb output should match. + + Returns: + string: The string that matched 'pattern', or an empty string if + 'pattern' was None. + """ + + match = '' + + gdb.sendline(command) + + if pattern: + # PExpect does a non-greedy match for '+' and '*'. Since it can't look + # ahead on the gdb output stream, if 'pattern' ends with a '+' or a '*' + # we may end up matching only part of the required output. + # To avoid this, we'll consume 'pattern' and anything that follows it + # up to and including the gdb prompt, then extract 'pattern' later. + index = gdb.expect([u'{0}.+{1}'.format(pattern, gdb_prompt), + pexpect.TIMEOUT]) + + if index == 0: + # gdb.after now contains the whole match. Extract the text that + # matches 'pattern'. + match = re.match(pattern, gdb.after, re.DOTALL).group() + elif index == 1: + # We got a timeout exception. Print information on what caused it + # and bail out. + error = ('Response does not match the expected pattern.\n' + 'Command: {0}\n' + 'Expected pattern: {1}\n' + 'Response: {2}'.format(command, pattern, gdb.before)) + + raise pexpect.TIMEOUT(error) + else: + # Consume just the the gdb prompt. + gdb.expect(gdb_prompt) + + return match + +def init_test(test_bin, printer_files, printer_names): + """Loads the test binary file and the required pretty printers to gdb. + + Args: + test_bin (string): The name of the test binary file. + pretty_printers (list of strings): A list with the names of the pretty + printer files. + """ + + # Load all the pretty printer files. We're assuming these are safe. + for printer_file in printer_files: + test('source {0}'.format(printer_file)) + + # Disable all the pretty printers. + test('disable pretty-printer', r'0 of [0-9]+ printers enabled') + + # Enable only the required printers. + for printer in printer_names: + test('enable pretty-printer {0}'.format(printer), + r'[1-9][0-9]* of [1-9]+ printers enabled') + + # Finally, load the test binary. + test('file {0}'.format(test_bin)) + +def go_to_main(): + """Executes a gdb 'start' command, which takes us to main.""" + + test('start', r'main') + +def get_line_number(file_name, string): + """Returns the number of the line in which 'string' appears within a file. + + Args: + file_name (string): The name of the file we'll search through. + string (string): The string we'll look for. + + Returns: + int: The number of the line in which 'string' appears, starting from 1. + """ + number = -1 + + with open(file_name) as src_file: + for i, line in enumerate(src_file): + if string in line: + number = i + 1 + break + + if number == -1: + raise NoLineError(file_name, string) + + return number + +def break_at(file_name, string, temporary=True, thread=None): + """Places a breakpoint on the first line in 'file_name' containing 'string'. + + 'string' is usually a comment like "Stop here". Notice this may fail unless + the comment is placed inline next to actual code, e.g.: + + ... + /* Stop here */ + ... + + may fail, while: + + ... + some_func(); /* Stop here */ + ... + + will succeed. + + If 'thread' isn't None, the breakpoint will be set for all the threads. + Otherwise, it'll be set only for 'thread'. + + Args: + file_name (string): The name of the file we'll place the breakpoint in. + string (string): A string we'll look for inside the file. + We'll place a breakpoint on the line which contains it. + temporary (bool): Whether the breakpoint should be automatically deleted + after we reach it. + thread (int): The number of the thread we'll place the breakpoint for, + as seen by gdb. If specified, it should be greater than zero. + """ + + if not thread: + thread_str = '' + else: + thread_str = 'thread {0}'.format(thread) + + if temporary: + command = 'tbreak' + break_type = 'Temporary breakpoint' + else: + command = 'break' + break_type = 'Breakpoint' + + line_number = str(get_line_number(file_name, string)) + + test('{0} {1}:{2} {3}'.format(command, file_name, line_number, thread_str), + r'{0} [0-9]+ at 0x[a-f0-9]+: file {1}, line {2}\.'.format(break_type, + file_name, + line_number)) + +def continue_cmd(thread=None): + """Executes a gdb 'continue' command. + + If 'thread' isn't None, the command will be applied to all the threads. + Otherwise, it'll be applied only to 'thread'. + + Args: + thread (int): The number of the thread we'll apply the command to, + as seen by gdb. If specified, it should be greater than zero. + """ + + if not thread: + command = 'continue' + else: + command = 'thread apply {0} continue'.format(thread) + + test(command) + +def next_cmd(count=1, thread=None): + """Executes a gdb 'next' command. + + If 'thread' isn't None, the command will be applied to all the threads. + Otherwise, it'll be applied only to 'thread'. + + Args: + count (int): The 'count' argument of the 'next' command. + thread (int): The number of the thread we'll apply the command to, + as seen by gdb. If specified, it should be greater than zero. + """ + + if not thread: + command = 'next' + else: + command = 'thread apply {0} next' + + test('{0} {1}'.format(command, count)) + +def select_thread(thread): + """Selects the thread indicated by 'thread'. + + Args: + thread (int): The number of the thread we'll switch to, as seen by gdb. + This should be greater than zero. + """ + + if thread > 0: + test('thread {0}'.format(thread)) + +def get_current_thread_lwpid(): + """Gets the current thread's Lightweight Process ID. + + Returns: + string: The current thread's LWP ID. + """ + + # It's easier to get the LWP ID through the Python API than the gdb CLI. + command = 'python print(gdb.selected_thread().ptid[1])' + + return test(command, r'[0-9]+') + +def set_scheduler_locking(mode): + """Executes the gdb 'set scheduler-locking' command. + + Args: + mode (bool): Whether the scheduler locking mode should be 'on'. + """ + modes = { + True: 'on', + False: 'off' + } + + test('set scheduler-locking {0}'.format(modes[mode])) + +def test_printer(var, to_string, children=None, is_ptr=True): + """ Tests the output of a pretty printer. + + For a variable called 'var', this tests whether its associated printer + outputs the expected 'to_string' and children (if any). + + Args: + var (string): The name of the variable we'll print. + to_string (raw string): The expected output of the printer's 'to_string' + method. + children (map {raw string->raw string}): A map with the expected output + of the printer's children' method. + is_ptr (bool): Whether 'var' is a pointer, and thus should be + dereferenced. + """ + + if is_ptr: + var = '*{0}'.format(var) + + test('print {0}'.format(var), to_string) + + if children: + for name, value in children.items(): + # Children are shown as 'name = value'. + test('print {0}'.format(var), r'{0} = {1}'.format(name, value)) + +def check_debug_symbol(symbol): + """ Tests whether a given debugging symbol exists. + + If the symbol doesn't exist, raises a DebugError. + + Args: + symbol (string): The symbol we're going to check for. + """ + + try: + test('ptype {0}'.format(symbol), r'type = {0}'.format(symbol)) + + except pexpect.TIMEOUT: + # The symbol doesn't exist. + raise DebugError(symbol) diff --git a/scripts/test_printers_exceptions.py b/scripts/test_printers_exceptions.py new file mode 100644 index 0000000000..17034b544e --- /dev/null +++ b/scripts/test_printers_exceptions.py @@ -0,0 +1,61 @@ +# Exception classes used when testing the Python pretty printers. +# +# Copyright (C) 2016 Free Software Foundation, Inc. +# This file is part of the GNU C Library. +# +# The GNU C Library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2.1 of the License, or (at your option) any later version. +# +# The GNU C Library 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 +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with the GNU C Library; if not, see +# <http://www.gnu.org/licenses/>. + +class NoLineError(Exception): + """Custom exception to indicate that a test file doesn't contain + the requested string. + """ + + def __init__(self, file_name, string): + """Constructor. + + Args: + file_name (string): The name of the test file. + string (string): The string that was requested. + """ + + super(NoLineError, self).__init__() + self.file_name = file_name + self.string = string + + def __str__(self): + """Shows a readable representation of the exception.""" + + return ('File {0} has no line containing the following string: {1}' + .format(self.file_name, self.string)) + +class DebugError(Exception): + """Custom exception to indicate that a required debugging symbol is missing. + """ + + def __init__(self, symbol): + """Constructor. + + Args: + symbol (string): The name of the entity whose debug info is missing. + """ + + super(DebugError, self).__init__() + self.symbol = symbol + + def __str__(self): + """Shows a readable representation of the exception.""" + + return ('The required debugging information for {0} is missing.' + .format(self.symbol)) |