Jim Warner [Wed, 29 Nov 2017 06:00:00 +0000 (00:00 -0600)]
library: correct 'use' calculations for <slabinfo> api
Awhile back, the calculation for cache utilization was
corrected to avoid a 32 bit overflow. This commit just
brings this branch into line with that earlier change.
[ references shown below represent the master branch ]
Jim Warner [Sun, 19 Nov 2017 06:00:00 +0000 (00:00 -0600)]
library: delete some obsolete parameter checking logic
This commit removes some obsolete parameter validation
code which was needed back when certain functions were
public, called directly by users (1st/2nd generation).
Now that they're static they can be safely eliminated.
Signed-off-by: Jim Warner <james.warner@comcast.net>
With older library logic having been modified to avoid
using those potentially deadly alloc.h routines, while
improving 'errno' handling, we're ready to standardize
and enhance newlib's approach to any potential errors.
In so doing, we'll establish the following objectives:
. . . . . . . . . . . . . functions returning an 'int'
. an error will be indicated by a negative number that
is always the inverse of some well known errno.h value
. . . . . . . . . . . functions returning an 'address'
. any error will be indicated by a NULL return pointer
with the actual reason found in the formal errno value
And, when errno is manipulated directly we will strive
to do so whenever possible within those routines which
have been declared with PROCPS_EXPORT. In other words,
in the user callable functions defined in source last.
[ But, that won't always be possible. In particular, ]
[ all the 'read_failed' functions will sometimes set ]
[ 'errno' so that they can serve callers returning a ]
[ NULL or an int without duplicating a lot of logic. ]
[ Also, that includes one subordinate function which ]
[ was called by 'read_failed' in the <slabinfo> API. ]
------------------------------------------------------
Along the way, several additional miscellaneous issues
were addressed. They're listed here now for posterity.
. the '-1' return value passed outside the library was
eliminated since it would erroneously equate to -EPERM
. the stacks_fetch functions in <diskstats> and <stat>
weren't checked for their possible minus return values
. hash create was not checked in <meminfo> or <vmstat>
. fixed 'new' function faulty parm check in <slabinfo>
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Wed, 15 Nov 2017 05:00:00 +0000 (00:00 -0500)]
library: eliminate all dependencies on alloc.h/alloc.c
While that old master branch library may utilize those
memory allocation functions found in the alloc module,
it was inappropriate for this newlib branch to subject
callers to a stderr message followed by an early exit.
Of course, the old libprocps offered a message handler
override provision (xalloc_err_handler) but that, too,
would seem to be inappropriate for our modern library.
[ remember the battles fought with that damn libnuma ]
So, this commit will tweak those old inherited sources
setting the stage for standardized return values/errno
settings in connection with a memory allocation error.
------------------------------------------------------
Along the way, we'll address the following miscellany:
. Completely eliminate usage of anything from alloc.h.
This, of course, entails our own error checking of the
alternative allocation calls from stdlib.h & string.h.
. Eliminate use of the strdup function where possible,
as with 'procps_uptime' and 'procps_loadavg' routines.
. Whack some obsolete code (getslabinfo) in sysinfo.c.
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Mon, 13 Nov 2017 05:00:00 +0000 (00:00 -0500)]
top: add and/or expand a couple more translation hints
After noticing that the 'uk' translation expanded what
was supposed to be a 3 line header into 5 lines, seems
appropriate to offer more guidance on max lines count.
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Sun, 8 Oct 2017 05:00:00 +0000 (00:00 -0500)]
top: stop neglecting potential utf8 field descriptions
And I thought those strange characters I saw with only
certain translations in Fields Management descriptions
were resulting from my terminal emulator deficiencies.
Turns out that ol' top wasn't addressing possibilities
of such descriptions ending with multi-byte sequences.
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Sat, 7 Oct 2017 05:00:00 +0000 (00:00 -0500)]
top: eliminate that potential vulnerability for TOCTOU
Initially, I was going to ignore that coverity warning
CID #177876. But, since top may be running SETUID it's
best if it can be avoided instead. The fix was simple.
We'll trade the access() call for a real fopen() call.
This time-of-check-time-of-use warning should go away.
------------------------------------------------------
When XDG support was originally introduced in top, the
author made a poor choice in access(). A real question
that needed asking was 'does the file exist'. However,
the question that was asked was 'can this real user ID
or this real group ID access the file'. Then, when the
fopen() is finally issued, top would use the effective
user ID or the effective group ID to access that file.
That's what opened the potential TOCTOU vulnerability,
which was important only if top was running SUID/SGID.
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Fri, 6 Oct 2017 05:00:00 +0000 (00:00 -0500)]
top: ensure bug report suggestion agrees with man page
The top man page was changed back on 10/20/15, in that
commit shown below. There, freelists.org was suggested
as the bug reports recipient. But, the program was not
changed from the original Debian bug reports approach.
Jim Warner [Wed, 4 Oct 2017 05:00:00 +0000 (00:00 -0500)]
top: make the 'utf8_proper_col' routine more efficient
This patch better exploits short-circuit evaluation in
two 'if' tests. In every case, the 1st of 2 conditions
in each 'if' test must take place but it always proves
true with each iteration for 1 of the 'if' statements.
Thus, the 2nd condition will have to be evaluated too.
By reordering 2 tests in each 'if', we can ensure that
the 2nd condition will then be tested much less often.
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Wed, 4 Oct 2017 05:00:00 +0000 (00:00 -0500)]
top: make that 'make_str_utf8' function more efficient
Upon reflection, there was absolutely no justification
for that call to strlen() which was then followed by a
call to snprintf(). The latter provides this needed #.
[ also make that 'delta' value a little more visible ]
[ instead of hiding it at the end of a its code line ]
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Tue, 3 Oct 2017 05:00:00 +0000 (00:00 -0500)]
top: ensured one translation hint agrees with template
Now that top can properly handle translated multi-byte
strings I've been reviewing translated efforts so far,
and weighing output against related translation hints.
In one case, a translation hint has not kept pace with
the current program state. In addition, that same hint
could be expanded to suggest translation alternatives.
[ frankly I never expected the translators to tackle ]
[ some of those 'special' strings. the task appeared ]
[ just too daunting. but they have done a great job! ]
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Tue, 3 Oct 2017 05:00:00 +0000 (00:00 -0500)]
top: again avoid multiple evaluation of macro argument
Before top was modified to exploit the new <pids> api,
there was protection in that task_show() makeVAR macro
to avoid multiple evaluation of this macro's argument.
But, in that commit referenced below, such a safeguard
was lost. This commit simply restores proper behavior.
Jim Warner [Tue, 3 Oct 2017 05:00:00 +0000 (00:00 -0500)]
top: when did scale_mem parm lose 'unsigned' qualifier
Well, for some strange unknown reason it happened in a
commit referenced below. But this patch reverts it and
puts this newlib scale_mem on par with the master guy.
[ a little more research reveals that it should have ]
[ been reverted in the 2nd commit shown. that's when ]
[ types were fixed after XTRA_PROCPS_DEBUG was used. ]
Jim Warner [Mon, 2 Oct 2017 05:00:00 +0000 (00:00 -0500)]
top: correct an insidious occasional truncation buglet
With the help of our Swedish translator, hopefully the
final buglet has now been vanquished in the multi-byte
translation support. This one was a real nasty bugger.
Although it didn't occur with every terminal emulator,
occasionally random text lines were being chopped off.
As it turns out, those terminals were blameless. There
were two separate places in top's show_special routine
where potential multi-byte sequences were inadequately
addressed. Solution: exploit existing utf-8 functions.
[ it also became apparent that the translation hints ]
[ in the top_nls module were deficient. so a special ]
[ caution was added regarding the final line of txt. ]
Jim Warner [Sun, 1 Oct 2017 22:59:59 +0000 (17:59 -0500)]
top: plug a small potential hole in multi-byte support
Unlike the insp_mkrow_raw function the insp_mkrow_utf8
routine is not equipped to print non-ctl, non-printing
characters like '<7f>'. However, technically that very
value currently slips through the cracks. So with this
patch top will now print a space in the unlikely event
a character with the value of 127 is ever encountered.
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Sun, 1 Oct 2017 16:59:59 +0000 (11:59 -0500)]
ps: don't use '+' truncation indicator with multi-byte
The ps program generally supports multi-byte sequences
in strings representing user and group names. However,
should a multi-byte sequence span the maximum width of
a column, the '+' inserted by ps to signify truncation
will corrupt that sequence, misaligning the text line.
Unfortunately, there's insufficient info returned from
the escape_str function (who calls escape_str_utf8) to
provide a robust response. So, this commit will revert
to the old standby of displaying a number when the '+'
character would've corrupted that multi-byte sequence.
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Sat, 30 Sep 2017 05:22:22 +0000 (00:22 -0500)]
top: extend utf-8 multi-byte support to users & groups
Since all the necessary utf-8 plumbing is now in place
this commit will extend multi-byte support to user and
group names. Now top will be on a par with the ps guy.
[ plus, it's also my way of showing appreciation for ]
[ all those investments silently made by translators ]
Jim Warner [Sat, 30 Sep 2017 05:11:11 +0000 (00:11 -0500)]
top: some minor tweaks to the utf-8 multi-byte support
Translatable column headers are supposed to be limited
to no more than 7 characters, even though some columns
are wider than that or even variable width. That value
of 7 is dictated by the Fields Management screen which
will otherwise truncate a column header longer than 7.
Our new utf-8 support did not adequately deal with the
potential need for truncation of column headers should
that limit of 7 be exceeded. This patch corrects that.
[ a few comments were also tweaked just a little bit ]
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Thu, 28 Sep 2017 05:33:33 +0000 (00:33 -0500)]
top: extend multi-byte support to 'Inspection' feature
The previous commit implemented multi-byte support for
the basic top user interaction and display provisions.
This commit completes multi-byte support by addressing
that 'Inspect Other Output' feature (the 'Y' command).
Few people probably exploit this very powerful feature
which allows the perusing of any file or piped output.
And even if nobody uses 'Y', someone will stumble over
it on the help screen and try it out. Assuming top was
not built with INSP_OFFDEMO defined, they'll end up on
the screen our translators have faithfully translated.
Without this patch, such a screen would display with a
bunch of 'unprintable' characters which will then show
in the standard (less-like) way as: '^A', '<C3>', etc.
In other words, those poor screens will be a big mess!
[ this program can even display an executable binary ]
[ while at that same time supporting Find/Find Next. ]
[ imagine, a file with no guarantee of real strings! ]
[ just try a Find using less with such binary files. ]
With this commit, the translated 'Y' demo screens will
now be properly shown, providing no invalid multi-byte
characters have been detected. Should that be the case
then they'll be displayed in that less-like way above.
And, if users go on to fully exploit this 'Y' command,
there is a good chance that a file or pipe might yield
output in a utf-8 multi-byte form. Should that be true
such output will thus be handled appropriately by top.
[ in many respects, this change was more challenging ]
[ than the basic support within the previous commit. ]
[ story of my life: least used = most effort needed. ]
Many thanks to our procps-ng translators which enabled
a proper test of these changed 'Y' command provisions:
. Vietnamese: Trần Ngọc Quân
. Polish: Jakub Bogusz
. German: Mario Blättermann
. French: Frédéric Marchal, Stéphane Aulery
[ and my sincerest apologies too, for my negligence! ]
Jim Warner [Thu, 28 Sep 2017 05:22:22 +0000 (00:22 -0500)]
top: refactored for correct multi-byte string handling
When this project first began implementing translation
support nearly 6 years ago, we overcame many 'gettext'
obstacles and limitations. And, of course, there were
not any actual translations at the time so our testing
was quite limited plus, in many cases, only simulated.
None of that, however, can justify or excuse the total
lack of attention to top's approach to NLS, especially
since some actual translations have existed for years.
When the issue referenced below was raised, I suffered
immediate feelings of anxiety, doubt and pending doom.
This was mostly because top strives to avoid line wrap
at all costs and that did not bode well for multi-byte
translated strings, using several bytes per character.
I was also concerned over possible performance impact,
assuming it was even possible to properly handle utf8.
But, after wrestling with the problem for several days
those initial feelings have now been replaced by guilt
over any trouble I initially caused those translators.
One can only imagine how frustrating it must have been
after the translation effort to then see top display a
misaligned column header and fields management page or
truncated screens like those of help or color mapping.
------------------------------------------------------
Ok, with that off my chest let's review these changes,
now that top properly handles UTF8 multi-byte strings.
. Performance - virtually all of this newly added cost
for multi-byte support is incurred during interactions
with the user. So, performance is not really an issue.
The one occasion when performance is impacted is found
during 'summary_show()' processing, due to an addition
of one new call to 'utf8_delta()' in 'show_special()'.
. Extra Wide Characters - I have not yet and may never
figure out a way to support languages like zh_CN where
the characters can be wider than most other languages.
. Translated User Name - at some future point we could
implement translation of user names. But as the author
of the issue acknowledged such names are non-standard.
Thus task display still incurs no new multi-byte costs
beyond those already incurred in that escape.c module.
For raising the issue I extend my sincerest thanks to:
Göran Uddeborg
Jim Warner [Thu, 28 Sep 2017 05:11:11 +0000 (00:11 -0500)]
top: correct a minor instance of wrong NLS macro usage
The 'N_fmt' and 'N_txt' macros are interchangeable and
just highlight the 2 str types found in Norm_nlstable.
The change in this patch (strictly cosmetic) was found
during the coding for what will be the next 2 commits.
It has not been squashed into either of those so as to
not muddy up the waters for what was a major refactor.
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Thu, 7 Sep 2017 05:00:00 +0000 (00:00 -0500)]
library: add delta values with swap too, <meminfo> api
As an oversight, delta values for SWAP amounts weren't
included in the <meminfo> API. Since any runtime costs
of including them only amount to slightly more storage
this commit will simply correct the earlier oversight.
Signed-off-by: Jim Warner <james.warner@comcast.net>
Craig Small [Sun, 24 Sep 2017 23:09:57 +0000 (09:09 +1000)]
vmstat: Fix alignment for disk partition format
The disk partition format translation hint, the actual text and the
printf statements were all diagreeing with each other. This meant
the disk partition output from vmstat was a mess.
Now all the titles and values line up and everything is right-aligned
which looks better than some half-hearted attempt at center-aligned
for titles and right-aligned for values.
When supplying the -p command to uptime, it does not display any
sections where the value is less than 1; however, after a reboot, this
causes the command to just output "up". Showing 0 minutes when the
system has been up for less than a minute makes it clear a reboot just
occurred.
Werner Fink [Fri, 7 Jul 2017 12:09:11 +0000 (14:09 +0200)]
Increase standard I/O buffer a lot to be able to
read huge informations at once as otherwise all files below
/proc/sys, not using the seq_file API at the kernel side,
will return EOF on a second read.
Jim Warner [Thu, 17 Aug 2017 07:22:22 +0000 (02:22 -0500)]
top: protect against the anomalous 'Mem' graph display
Until this patch, top falsely assumed that there would
always be some (small) amount of physical memory after
subtracting 'used' and 'available' from the total. But
as the issue referenced below attests, a sum of 'used'
and 'available' might exceed that total memory amount.
I'm not sure if this is a problem with our calculation
of the 'used' amount, a flaw in the kernel 'available'
algorithms or some other reason I cannot even imagine.
Anyway, this patch protects against such a contingency
through the following single line addition of new code
. if (pct_used + pct_misc > 100.0 || pct_misc < 0) ...
The check for less than zero is not actually necessary
as long as the source numbers remain unsigned. However
should they ever become signed, we'll have protection.
[ Most of the changes in this commit simply separate ]
[ a variable's definition from its associated logic. ]
Jan Rybar [Sat, 5 Aug 2017 22:00:00 +0000 (00:00 +0200)]
top: refresh interval accepts non-locale decimal value
For the past 3 years top has fully honored that locale
LC_NUMERIC setting which impacts his refresh interval.
For the past nearly 5 years top has saved that refresh
value in a locale independent form in his config file.
With this commit we'll intentionally break top so that
a comma or period will be accepted for the radix point
regardless of what that LC_NUMERIC may have suggested.
The current locale LC_NUMERIC will, however, determine
how the delay interval is displayed in the 'd' prompt.
[ This position is better than the approach employed ]
[ by those coreutils 'sleep' and 'timeout' programs. ]
[ Both claim to permit floating point arguments. But ]
[ neither one will accept the comma separator should ]
[ the locale be a country that in fact uses a comma. ]
Jim Warner [Sun, 30 Jul 2017 05:00:00 +0000 (00:00 -0500)]
top: at program startup, avoid a little extra overhead
There was a time when the PROCPS_PIDS_noop represented
the highest valued enumerator. Therefore if one wished
to prime an array consisting of pids_result structures
with this specific item a loop would have been needed.
Now that this enum is the first one, with the value of
zero, we can avoid avoid such a loop with just calloc.
But just in case, we'll use an 'if' to guarantee zero.
[ and the nice thing is, since the value is known at ]
[ compile time, that 'if' test plus subordinate loop ]
[ can be discarded by the compiler as long as it's 0 ]
Jim Warner [Wed, 21 Jun 2017 01:20:20 +0000 (20:20 -0500)]
top: fixing command line parsing errors is now a habit
Ok, I admit it. I'm now tired of cleaning up after me.
This is the 3rd related tweak after that '-1' argument
was originally introduced. And with this patch we will
once again properly honor the '-o' and '-u|U' switches
without a need to be followed by an additional switch.
[ one can follow my unfortunate trail of alterations ]
[ beginning with my most recent fix referenced below ]
Jim Warner [Tue, 13 Jun 2017 05:00:00 +0000 (00:00 -0500)]
library: correct the #define FALSE_THREADS, <PIDS> api
Awhile back, that former QUICK_THREADS #define evolved
into the development (only) FALSE_THREADS which can be
used to ensure a 'duplicate ENUM' convention is output
when certain string fields can't be easily duplicated.
Unfortunately, that original implementation was marred
with zeros being displayed for /proc/$$/meminfo fields
in all the child threads for a multi-threaded process.
So this commit corrects that zero memory field buglet.
Jim Warner [Mon, 29 May 2017 07:22:22 +0000 (02:22 -0500)]
misc: adapt 3 programs to some enum changes, <meminfo>
That preceding commit corrected a little mismanagement
regarding the MEMINFO_MEM_CACHED enumerator, which was
used in the following programs: free, top plus vmstat.
This patch simply adapts those programs to use the new
MEMINFO_MEM_CACHED_ALL enumerator, which reflects both
the 'Cached' plus 'SReclaimable' values they expected.
Signed-off-by: Jim Warner <james.warner@comcast.net>
The <meminfo> module attempted to duplicate the former
sysinfo memory calculations wherein 'SReclaimable' was
added to 'Cached' for the 'kb_main_cached' equivalent.
But, this original approach was wrong for two reasons.
1. The addition occurred too late to impact the 'USED'
calculation which could then cause an underflow in the
top memory display if 'SReclaimable' was heavily used.
2. In changing the actual /proc/meminfo 'Cached' value
it meant that users could not rely on that proc(5) man
page when interpreting the MEMINFO_MEM_CACHED results.
So this commit adds a new enumerator for the inclusive
cached amount plus repositions the calculation so that
a MEMINFO_MEM_USED result will exclude 'SReclaimable'.
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Mon, 22 May 2017 05:00:00 +0000 (00:00 -0500)]
top: correct functioning of the '-p' command line args
With the introduction of a new '1' command line toggle
I have gone and broken a provision of the '-p' command
line switch (pids monitoring). Multiple pids could not
be specified through the use of comma delimited lists.
Thus, this commit simply corrects that newly added bug
which was born in the 'adjustment' commit shown below.
Werner Fink [Mon, 22 May 2017 12:31:34 +0000 (22:31 +1000)]
sysctl: Print lines longer than 1024 chars
as well do not open /proc/sys files if only the names of the
system control names of the kernel parameters should be shown.
Avoid leaking tmpname in case of a pattern mismatch.
Craig Small [Mon, 22 May 2017 12:15:59 +0000 (22:15 +1000)]
kill: -l space between name parses correctly
This was supposed to be just a cherry-pick of the referenced
commit. However there were two problems:
1. kill code was moved out to its own file
2. strtosig() had a latent bug where signal numbers were not
converted to names.
Original note:
kill -lHUP would work correctly, but kill -l HUP would not.
The list option in kill was hit by a quirk of getopt_long where an
option with an optional argument would not attempt to get the argument
beyond the space, even though a mandatory argument would do that.
The fix is a kludge to scan to the next argument and if it looks
like something we can use, use it. Lucky for us, the list option is
one where parsing can stop immediately.
Jim Warner [Wed, 17 May 2017 05:00:00 +0000 (00:00 -0500)]
top: address the argument parsing quirk involving '-h'
There exists the possibility that a 'putp' call can be
issued before the 'setupterm' invocation has occurred,
as is reflected in a bugzilla report referenced below.
Strangely, such a SEGV isn't always triggered as logic
would suggest it ought to be. I experienced a fault in
these environments with the associated curses version:
. archlinux, procps-ng 3.3.12, ncurses 6.0.20170429
. fedora-25, procps-ng 3.3.10, ncurses 6.0.20160709
. opensuse-42.2, procps-ng 3.3.9, ncurses 5.9.20140201
. gentoo, procps-ng 3.3.12, ncurses 6.0.20150808
. slackw-14.2, procps-ng 3.3.12, ncurses 6.0.20160910
Whereas under these environments there was no problem:
. ubuntu-17.04, procps-ng 3.3.12, ncurses 6.0.20160625
. debian-test, procps-ng 3.3.12, ncurses 6.0.20161126
. mageia-5.1, procps-ng 3.3.9, ncurses 5.9.20140323
[ as an aside, the expected result in the bug report ]
[ is incorrect and should mention the '1' parameter. ]
[ however, until release 3.3.13 when the '1' becomes ]
[ a valid switch, numbers are not detected when used ]
[ with any switch which doesn't require an argument. ]
[ you're welcome to treat that as a separate bugglet ]
Jim Warner [Fri, 12 May 2017 05:03:00 +0000 (00:03 -0500)]
library: set stage for NUMA node field display support
In response to that suggestion referenced below, these
changes allow display of task/thread level NUMA nodes.
Currently, only the 'top' program offers any NUMA type
support and it is limited to the Summary Area display.
With this commit both the 'top' and 'ps' programs will
be able to display NUMA nodes associated with threads.
Jim Warner [Fri, 12 May 2017 05:02:00 +0000 (00:02 -0500)]
misc: eliminate all those remaining gcc -Wall warnings
Reference(s):
proc/readproc.c: In function 'statm2proc'
proc/readproc.c:600:9: warning: variable 'num' set but not used [-Wunused-but-set-variable]
proc/stat.c: In function 'stat_derive_unique':
proc/stat.c:429:1: warning: no return statement in function returning non-void [-Wreturn-type]
ps/parser.c: In function 'arg_type':
ps/parser.c:1098:3: warning: this 'if' clause does not guard... [-Wmisleading-indentation]
ps/parser.c:1099:34: note: ...this statement, but the latter is misleadingly indented as if it is guarded by the 'if'
lib/signals.c: In function 'strtosig':
lib/signals.c:243:9: warning: this 'if' clause does not guard... [-Wmisleading-indentation]
lib/signals.c:245:13: note: ...this statement, but the latter is misleadingly indented as if it is guarded by the 'if'
slabtop.c: In function 'print_summary':
slabtop.c:223:29: warning: unused variable 'stats' [-Wunused-variable]
watch.c: In function 'process_ansi':
watch.c:232:5: warning: this 'if' clause does not guard... [-Wmisleading-indentation]
watch.c:235:2: note: ...this statement, but the latter is misleadingly indented as if it is guarded by the 'if'
Signed-off-by: Jim Warner <james.warner@comcast.net>
When those standardized 'derived' TIC enumerators were
introduced, a problem with potential DELTA distortions
was also introduced when toggling cpus offline/online.
It has always been true that the 1st (summary) line in
/proc/stat will experience a decrease in total tics if
a new cpu is brought online. Such decreases are mostly
due to reductions in 'idle' and 'iowait' tics. Exactly
why such a counterintuitive phenomenon should occur is
a mystery, but this has been acknowledged in proc.txt.
A separate potential distortion arises with individual
cpus. And, here it extends to both bringing processors
online plus taking them offline too. When that happens
the order of the cpus array tracking is upset, placing
the 'new' values in some other processor's array slot.
But even if we were to occupy the same slot, the issue
regarding reductions in 'idle' & 'iowait' still apply.
In all cases, when a DELTA field was found to be minus
it was forced to zero via the 'TICsetH' macro. However
the 'derived' calculations are subject to new forms of
distortion with their own DELTA values. For example we
could find DELTA_SUM_USER + DELTA_SUM_SYSTEM exceeding
DELTA_SUM_TOTAL, an illogical/inappropriate condition.
So this commit moves former protections for individual
cpus to the stat_derive_unique() function and modifies
it to also extend protections to the 'derived' values.
In the process we now protect the cpu 'summary' counts
which were unfortunately previously overlooked (oops).
Jim Warner [Tue, 21 Mar 2017 05:00:00 +0000 (00:00 -0500)]
top: make command line switch parsing even more robust
This program has always tried to maintain an extermely
robust command line parsing procedure, far more robust
that what's available with the getopt stuff. But, with
the introduction of our first numeric switch it should
have been made even more robust than, in fact, it was.
This commit will now accomplish such a desirable goal.
Jim Warner [Mon, 13 Mar 2017 11:00:00 +0000 (06:00 -0500)]
top: by default, show cmd line vs. cmd name at startup
All of top's display was designed to fit into an 80x24
terminal. This includes the help screens plus both the
Summary and Task Areas, assuming no saved config file.
With release 3.3.10, the startup defaults were changed
assuming ./configure --disable-modern-top wasn't used.
This was done in the hope of introducing some users to
unknown capabilities such as colors, forest view, etc.
The purpose of this commit is to coax a few more users
into possibly exploring another capability: scrolling.
We do so by tweaking the default startup display so as
to show full command lines. Now, when things no longer
fit in 80x24, horizontal scrolling might be exploited.
[ of course, this can be reversed with the -c switch ]
[ this patch has been adapted from the master branch ]
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Mon, 13 Mar 2017 10:00:00 +0000 (05:00 -0500)]
top: provide -1 command line switch as 'Cpu(s)' toggle
If built without ./configure --disable-modern-top, the
program displays each cpu individually providing there
is sufficient vertical screen real estate. For massive
SMP environments this will necessitate use of a config
file where the cpu summary toggle ('1') could be saved
via the 'W' command. But, an rcfile may not be viable.
So this commit introduces a '1' command line switch to
emulate exactly the effects of the interactive toggle.
And since it is our first numeric switch some existing
parsing logic had to be changed slightly. Such changes
are, in truth, an improvement. For example, instead of
seeing "inappropriate '2'" with ./top -2 we'll now see
the vastly more appropriate error "unknown option '2'.
Jim Warner [Mon, 13 Mar 2017 09:00:00 +0000 (04:00 -0500)]
top: provide -E command line switch for memory scaling
In their 3.2.7 version of top, Redhat introduced an -M
switch to automatically scale Summary Area memory data
to avoid truncation (and the resulting '+' indicator).
The procps-ng top does not employ suffixes with memory
data nor does it allow for different scaling with each
separate value. Rather, scaling appears at line start.
If built without ./configure --disable-modern-top, the
Summary Area memory will be scaled at GiB which should
lessen chance of truncation. Otherwise KiB was used to
reflect such memory, increasing the truncation chance.
And while 'W' can be used to preserve some appropriate
scaling value, there are arguments against such rcfile
approaches as cited in the issue and bug report below.
So this commit will bump the Summary Area memory scale
factor from KiB to MiB when using --disable-modern-top
as a concession to that Redhat bug report noted below.
And it also introduces a new command line switch which
can force any desired scaling regardless of the rcfile
or which ./configure option might have been specified.
[ for top's help text we'll show 'E' as if it were a ]
[ switch without arguments in order to keep the help ]
[ text displayable without wrap in an 80x24 terminal ]
[ the man page, however, will show all k-e arguments ]
Jim Warner [Mon, 13 Mar 2017 08:00:00 +0000 (03:00 -0500)]
top: show fewer decimal places for memory (by default)
After much reflection I've come to the conclusion that
displaying 3 decimal places (usually) when memory data
had been scaled is no longer optimal with today's ever
increasing amounts. And given that not all task memory
fields are the same widths, inconsistencies can easily
arise as illustrated and discussed in the issue below.
Instead of unilaterally reducing the number of decimal
places, this commit will sneak in such a change via an
existing configure option that was very likely unused.
The former 'disable-wide-memory' option has now become
'enable-wide-memory', which can be used if the current
behavior (3 decimal places) is preferred. Without that
option, whenever memory is scaled beyond KiB, just one
decimal place will be shown in Summary and Task areas.
And Task area field width will no longer be changed by
this revised configure option. Instead, all such field
widths will now be fixed at the former maximum values.
Craig Small [Sat, 4 Feb 2017 00:44:01 +0000 (11:44 +1100)]
pkill: Return 0 if successfully killed process
Original report:
When trying kill a process with insufficient privileges (see blow),
pkill displays the error message “... failed: Operation not permitted”,
but returns 0. Surely it should return 3?
Return value 0 means one of more things matched. For a pgrep (which
shares code with pkill) this makes sense, there was a match. It seems
wrong for pkill to return 0 when it in fact could not do what you told
it to. However return value 3 means a fatal error and it's not fatal.
Looking at other programs when trying to kill things it cannot kill.
shell kill returns 1, procps kill returns 1, killall returns 1, skill
returns 0 (and says it was successful!, ah well poor old skill)
The consensus seems to be that you return 1 if you cannot kill it, even
if you found it. In other words the return value for both not found and
not able to kill it is the same.
pkill only returns 0 if something was killed. This means we found a
match AND the kill() system call worked too.
Jim Warner [Thu, 26 Jan 2017 06:00:00 +0000 (00:00 -0600)]
top: adapt the master branch XDG specification support
This patch adapts a master branch commit to our newlib
branch. Shown below was that original commit msg text.
------------------------------------------------------
top: Add unobtrusive XDG support
By default the file HOME/.toprc will be prefered. This ensures there
should be minimal breakage even if this file is later created by some
other means. Otherwise we will follow the new behaviour described by
the XDG Base Directory Specification:
If the XDG_CONFIG_HOME environment variable is available we will attempt
to use this as XDG_CONFIG_HOME/procps/toprc otherwise we will fall-back
to HOME/.config/procps/toprc instead.
Jim Warner [Sat, 31 Dec 2016 06:00:00 +0000 (00:00 -0600)]
library <stat>: input file buffer size must be dynamic
Since its introduction, our evolved /proc/stat API has
relied on a static buffer of 8192 bytes. This approach
is probably Ok for other /proc files but it would only
accommodate around 100 processors. If such a threshold
were exceeded then this interface could never succeed.
Now days 100 processors doesn't seem at all excessive.
So this commit trades that static buffer for a dynamic
self-tuning one. And since so much former top CPU code
was already rolled into this module, we just stole the
already proven top dynamic buffer management code too.
[ this also meant switching low level unbuffered I/O ]
[ calls to standard library buffered I/O calls. that ]
[ is exactly what <slabinfo> and <diskstats> employ. ]
Jim Warner [Sat, 31 Dec 2016 06:00:00 +0000 (00:00 -0600)]
library <stat>: improve response to cpu offline/online
With the addition of those new derived SUM values, any
CPUs taken offline or brought online would distort the
historical (delta) results. So this patch just forces
a history reset when such transitions are encountered.
Jim Warner [Sat, 31 Dec 2016 06:00:00 +0000 (00:00 -0600)]
library <slabinfo>: make read function name consistent
For each of those interfaces employing a priming read,
all the other 'read' functions begin with the module's
name except this guy which began with 'read_slabinfo'.
Now, they'll all begin with their module name then end
the same with a '_read_failed' boolean hinting suffix.
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Sat, 31 Dec 2016 06:00:00 +0000 (00:00 -0600)]
library: eliminate distorted history 1st time switches
Upon reflection, at the point where the 'priming read'
was introduced, any possibility of history distortions
was also eliminated. This was true because all of the
'old' (zeroed) data will have been replaced with 'new'
data whenever a user finally calls get, select & reap.
Thus, any DELTA values will automatically reflect that
interval between 'new' and subsequent retrieval calls.
[ diskstats didn't actually employ a 1st time switch ]
[ like the others so we have changed a comment only. ]
[ but that module will retain something similar used ]
[ inside node_update whenever a new node is created. ]
Jim Warner [Tue, 27 Dec 2016 14:08:08 +0000 (08:08 -0600)]
top: exploits several <stat> new category calculations
This commit just exploits those new library provisions
for tic categories, introduced in the preceding patch,
which had been prompted by the issue referenced below.
[ ok it also corrects the top graph for system usage ]
[ since this turkey failed to include tics for these ]
[ two interrupts: STAT_TIC_IRQ and STAT_TIC_SOFTIRQ. ]
Jim Warner [Tue, 27 Dec 2016 14:08:08 +0000 (08:08 -0600)]
library <stat>: standardized new category calculations
This commit arose out of the discussion (and research)
surrounding the issue cited below. It is an attempt to
consolidate and standardize the calculation of jiffies
categories (e.g. 'idle', 'busy', etc.) once & for all.
Also included is the enum STAT_TIC_NUM_CONTRIBUTORS in
case anyone, in the future, decides to calculate usage
based upon elapsed time * Hz (like top does in process
level %CPU stats). In such an event, a total number of
CPUs or NUMA Nodes would be needed for proper scaling.
Unlike the ps kludge under the master branch to ensure
that namespaces appear the same under both 32 & 64-bit
models, this newlib branch already used a proper type.
However source data still carried the original type as
'signed long' versus that more proper 'unsigned long'.
So, this patch makes sources & destinations identical.
Jim Warner [Sat, 3 Dec 2016 09:03:03 +0000 (03:03 -0600)]
ps: now made responsive to 32 or 64-bit address widths
While a Debian bug report referenced below was limited
to the 'eip' and 'esp' fields, this patch also extends
address width adaptations to some other addresses too.
[ and, we do so in a far less invasive manner than a ]
[ redhat approach shown below adding two new fields! ]
Reference(s):
. new debian bug report
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=846361
. old redhat solution
https://bugzilla.redhat.com/show_bug.cgi?id=244152
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Tue, 22 Nov 2016 15:10:10 +0000 (10:10 -0500)]
library: protect against possible 'refcount' underflow
In each module employing a priming read at 'new' time,
should that read fail, a call to 'unref' will be made.
However, there is a hidden dependency that these calls
must never occur before the context 'refcount' was set
due to the way an 'unref' conditional was constructed.
So this commit just ensures that 'unref' will function
as expected, even if called with a 'refcount' of zero.
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Tue, 11 Oct 2016 05:00:00 +0000 (00:00 -0500)]
top: tweak some stuff relating to non-displayed fields
In the commit referenced below, in addition to several
tweaks to comments, 3 fields were no longer assured of
being present in the results stacks. However, 2 of the
3 fields might, in fact, be required even if they were
not currently being displayed in any of the 4 windows.
The PIDS_CMD is used in two separate 'Inspect' headers
('Y' command) and the PIDS_ID_EUID is required if that
'User Filter' ('u' or 'U' command) was being employed.
That latter field's inclusion will be made conditional
but the former field must be unconditionally included.
( for old top, PIDS_CMD would have always been there )
Jim Warner [Fri, 7 Oct 2016 10:55:55 +0000 (05:55 -0500)]
ps: respond to loss of that PIDS_WCHAN_ADDR enumerator
No longer will ps print nwchan as 'ffffff', '-' or '1'
since the proc/PID/stat wchan field didn't represent a
real address anyway. Rather, the field will henceforth
output a dash ('-'), the ps customary 'not available'.
That man document was also tweaked to better represent
actual behavior. An asterisk ('*') was never shown for
threaded tasks and that dash ('-') usually didn't mean
running tasks (sometimes associated with permissions).
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Fri, 7 Oct 2016 09:44:44 +0000 (04:44 -0500)]
library <stat>: remove that PIDS_WCHAN_ADDR enumerator
Removing the Item_table 'stat' oldflags for WCHAN_ADDR
was wrong since that 'stat' field is not a constant 0.
Rather, it could assume these 3 values: -1, 0, and +1.
I have not been able to pin down a '-1' result, but it
probably means some sort of permission error (-EPERM).
The '1' or '0' values were supposed to distinguish the
tasks that were or were not blocked (whether there was
a wchan address). However, in practice there is little
correlation between those values and availability of a
kernel symbol in /proc/$$/wchan (perhaps due to race).
Anyway, the real point is that a 'stat' wchan does not
now intentionally contain an address. Thus, outputting
'ffffff', '-' or '1' in programs like ps is senseless.
So this patch just eliminates PIDS_WCHAN_ADDR from our
item enumerators leaving only the PIDS_WCHAN_NAME guy.
Now the new library can't be blamed for bad addresses!
Jim Warner [Fri, 7 Oct 2016 07:22:22 +0000 (02:22 -0500)]
top: make that 'forest view' just a tad more efficient
It makes no sense to begin our tracked nested level at
'1' then later require a '1' to be subtracted from the
level as artwork and indentation is added for display.
By beginning such tracked levels at zero, we can avoid
the need to adjust it & use it directly in a snprintf.
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Wed, 28 Sep 2016 18:13:13 +0000 (13:13 -0500)]
top: adapt for loss of fields obsoleted with linux 2.6
It would have been nice to remove this 'nDRT' guy from
the fields management screen and man document as well.
But, while this version of top could be made to handle
an older rcfile, the reverse would not have been true.
Besides, it's been zero for so long already we'll just
include a 'deprecated' note in top's man page for now.
[ the 'nTH ' field number was corrected in there too ]
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Wed, 28 Sep 2016 16:11:11 +0000 (11:11 -0500)]
library <pids>: remove fields obsoleted with linux 2.6
It seems inappropriate to blindly include fields known
to always be zero in our brand new library. Therefore,
this patch removes support for three such enumerators.
[ that stat 'it_real_value' (PIDS_ALARM) field could ]
[ have been made obsolete before a linux 2.6 release ]
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Wed, 28 Sep 2016 15:10:10 +0000 (10:10 -0500)]
library: strictly cosmetic, absolutely no code changes
This commit just contains some tweaks to comments plus
a few adjustments to whitespace for alignment purposes
and a normalization of the header inclusion #define's.
[ plus a spelling error in one header file was fixed ]
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Wed, 21 Sep 2016 13:08:08 +0000 (08:08 -0500)]
library <stat>: added overlooked numa guest tic counts
When this module was upgraded to 3rd generation in the
patch referenced below, numa node support was migrated
from the top program into newlib. The 'guest_nice' and
'guest' tics were overlooked as top did not need them.
So, this commit corrects that oversight and achieves a
proper symmetry between the cpu & numa jiffies counts.
Jim Warner [Sat, 17 Sep 2016 09:44:44 +0000 (04:44 -0500)]
library: add priming read at 'new' time <most modules>
A priming read at 'new' time in that <slabinfo> module
was important so that permission problems are detected
early. Plus, it also had the potential of making delta
values valid when 'get' or 'select' were first called.
It is for that latter reason that such a read was also
incorporated in the <diskstats> module 'new' function.
No other module, however, employed such priming reads.
This patch just brings those potential benefits to all
of our other newlib modules with the exception of that
<pids> guy. That module is, of necessity, sufficiently
different from those others to justify such exclusion.
Not only are there precious few DELTA enums in <pids>,
but the costs of a priming read would be much greater.
[ otherwise, these newly added priming reads have no ]
[ measurable negative impact on performance/timings. ]
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Sat, 17 Sep 2016 08:33:33 +0000 (03:33 -0500)]
library: summary name now more descriptive, <slabinfo>
The <slabinfo> header provides 3 groups of enumerators
with prefixes of SLABINFO, SLABS & SLABNODE. The first
is strictly user oriented & isn't supported internally
by any structure. The other two, however, have structs
associated with 'em but, unfortunately, 1 is misnamed.
The 'struct slabs_node' is associated with 'nodes' and
supports the enumerators with the SLABNODE prefix. But
the 'struct slabs_hist' was associated with 'hist' yet
supports those enumerators with just the SLABS prefix.
We do not care very much what some structure is called
but we do care about an identifier used manipulate it.
This patch will trade the 'hist' identifier associated
with 'struct slabs_hist' for a more congruous 'slabs'.
[ it's awful when the author can't remember what the ]
[ true meaning of an identifier is after creating it ]
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Sat, 17 Sep 2016 07:22:22 +0000 (02:22 -0500)]
library: improve support of dynamic numa nodes, <stat>
If, in fact, numa nodes are dynamic (like that current
total of on-line cpus) the existing logic was lacking.
It included an early return before checking the total.
So, this commit ensures that the nodes total is always
set or updated consistently in only a single function.
There's no need to set it at the time 'new' is called.
[ and since under our existing code this nodes total ]
[ could never possibly have been zero, the erroneous ]
[ test (with the early return) has now been whacked! ]
Signed-off-by: Jim Warner <james.warner@comcast.net>
Jim Warner [Sat, 17 Sep 2016 06:11:11 +0000 (01:11 -0500)]
top: remove explicit references to NUMA_DISABLE define
Since our library is responsible for NUMA support, and
since the top program already accommodates the lack of
NUMA data, there's no reason that #define NUMA_DISABLE
need be explicitly referenced in the top source files.
Ergo, this commit just eliminates all such references.
Now, top will rely only on procps_stat_reap() results.
Signed-off-by: Jim Warner <james.warner@comcast.net>
projects / procps-ng / log