GSOC: Introduction

Tue Mar 14 03:14:14 UTC 2017

On Tue, Mar 14, 2017 at 2:00 AM, Dmitry V. Levin <ldv at altlinux.org> wrote:
> On Tue, Mar 14, 2017 at 12:40:46AM +0530, Abhishek Tiwari wrote:
>> > Per-file summary is not in GNU change log format, please refer to
>> > https://www.gnu.org/prep/standards/html_node/Style-of-Change-Logs.html .
>> > Specifically, it lacks asterisks at the beginning of each new file
>> > description and has excess spaces between file part and description,
>> > otherwise looks fine.
>> >
>> > The patch wires up {,l}stat{,64}, fstatat/fstatat64 syscalls (used for obtaining
>> > file status), ustat syscall (which is a deprecated way to get some FS
>> > statistics) and some statfs syscalls (which are more contemporary way
>> > obtaining FS statistics information). On the other side, at least
>> > fstat/fstat64, fstatfs/fstatfs64, old{,f,l}stat, osf_{,f}statfs, and various
>> > (mostly unsupported) syscalls with osf_, svr4_, sysv_, bsd43_, and posix_
>> > prefixes, present on alpha and mips, are omitted. I'm not sure whether
>> > it was intended.
>> >
>> > There are minor tabulation irregularities introduced (at least) for
>> > newfstatat and fstatat64 syscall entries, it is better to avoid this.
>>
>> I have fixed the commit entries.
>> Now the patch includes all the stat-like syscalls that can be
>> displayed by following command->
>
> The idea is not to throw all syscalls that do some kind of stat into a
> single class, but to create a class of syscalls that could be useful.
> As Eugene already noted, there is little use of grouping syscalls that
> stat files with those that stat filesystems.
>
> Actually, there is a software project that could benefit from more
> fine-grained classes, e.g. oldstat+stat+stat64, oldlstat+lstat+lstat64,
> and newfstatat+fstatat64.

Ok, so oldstat+stat+stat64 is one class that is to be grouped under
%stat, oldstat+lstat+lstat64 as another class and
fstatat+newfstatat+fstatat64 and some other class.

I read in mailing list it is intended to group syscalls that perform
same action across different versions.

But it is little confusing to decide what to group into one class.

> There may be some use of a class that groups all filesystem stat syscalls
> like ustat and *statfs*.  This might appear to be easier for you to
> implement, especially the part that tests it.

I would be doing this part first. Grouping ustat and *statfs* under
%statfs. Is this class name fine?

>> grep -nr stat linux/*/syscallent*
>
> btw, why do you pass -nr to this grep command?

This grep command is just to display the places where I have made the
change. The command actually used is mentioned in commit.
>> >> --- a/syscall.c
>> >> +++ b/syscall.c
>> >> @@ -77,6 +77,7 @@
>> >>  #define TS TRACE_SIGNAL
>> >>  #define TM TRACE_MEMORY
>> >>  #define TSC TRACE_SCHED
>> >> +#define TST  TRACE_STAT
>> > Tabulation is used for separating macro declaration and definition, but
>> > neighbouring lines utilise space for this purpose.
>>
>> Please elaborate, i don't get what correction is intended here?
>
> The rule of thumb is simple:
> when patching a piece of code, follow its coding style.
>
> If the code uses spaces, use spaces.  If it uses tabs, use tabs.
> strace has a long history with a lot of contributors so the style
> used in different parts might differ slightly.

Sorry, that space was not intensional.

>> >> +++ b/tests/stat_like.test
>> >> @@ -0,0 +1,58 @@
>> >> +#!/bin/sh
>> >> +
>> >> +# Check how stat-like syscalls are traced.
>> >> +#
>> >> +# Copyright (c) 2017 The strace developers.
>> >> +# All rights reserved.
>> >> +#
>> >> +# Redistribution and use in source and binary forms, with or without
>> >> +# modification, are permitted provided that the following conditions
>> >> +# are met:
>> >> +# 1. Redistributions of source code must retain the above copyright
>> >> +#    notice, this list of conditions and the following disclaimer.
>> >> +# 2. Redistributions in binary form must reproduce the above copyright
>> >> +#    notice, this list of conditions and the following disclaimer in the
>> >> +#    documentation and/or other materials provided with the distribution.
>> >> +# 3. The name of the author may not be used to endorse or promote products
>> >> +#    derived from this software without specific prior written permission.
>> >> +#
>> >> +# THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
>> >> +# IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
>> >> +# OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
>> >> +# IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
>> >> +# INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
>> >> +# NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
>> >> +# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
>> >> +# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
>> >> +# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
>> >> +# THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
>> >> +
>> >> +STAT_TESTS='21 fstat
>> >> +33 ustat
>> >> +32 statx
>> >> +17 statfs
>> >> +21 oldfstat'
>> >> +
>> >> +NON_STAT_TESTS='11 fchdir
>> >> +27 futex
>> >> +10 fsync'
>> >> +
>> >> +. "${srcdir=.}/init.sh"
>> >> +
>> >> +echo "$STAT_TESTS" | while read w i
>> >> +do
>> >> +     run_prog ./$i > /dev/null
>> >> +     run_strace -a$w -e%stat ./$i > "$EXP"
>> >> +     match_diff "$LOG" "$EXP"
>> >> +done
>> >> +
>> >> +echo '+++ exited with 0 +++' > "$EXP"
>> >> +
>> >> +echo "$NON_STAT_TESTS" | while read w i
>> >> +do
>> >> +     run_prog ./$i > /dev/null
>> >> +     run_strace -a$w -e%stat ./$i > /dev/null
>> >> +     match_diff "$LOG" "$EXP"
>> >> +done
>> >> +
>> >> +rm "$EXP"
>> > This test produces the following diagnostics:
>> >
>> >     1,2c1,2
>> >     < fstat(0, 0x7f4dc9b2ffe0) = -1 EFAULT (Bad address)
>> >     < fstat(0, {st_dev=makedev(9, 1), st_ino=8761399, st_mode=S_IFREG|0640, st_nlink=1, st_uid=1000, st_gid=1000, st_blksize=4096, st_blocks=0, st_size=43147718418, st_atime=1969-12-31T21:59:17+0100.000000135, st_mtime=1969-12-31T21:59:19+0100.000000246, st_ctime=2017-03-13T07:03:55+0100.962268762}) = 0
>> >     ---
>> >     > stat("/etc/localtime", {st_mode=S_IFREG|0644, st_size=2246, ...}) = 0
>> >     > stat("/etc/localtime", {st_mode=S_IFREG|0644, st_size=2246, ...}) = 0
>> >     stat_like.test: failed test: ../strace -a21 -e%stat ./fstat output mismatch
>> >
>> > It exits with status 0, however.
>>
>> Please help me with this part, how can I run this test on my machine.
>
> Do you ask about running this particular test?  I think it's as simple as
> $ make check TESTS=stat_like VERBOSE=1

Thank you, I just never did testing this way and I am new to this part.
>
> --
> ldv
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> Strace-devel mailing list
> Strace-devel at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/strace-devel
>