Provide symbolic printing of some keywords by capitalizing them.

We only need -Wno-format-y2k for print.c .

branches: 1.29.44;
Document non-literal formats

g/c procfs fallback code, it's redundant now that we use sysctl to get
process info

branches: 1.27.6;
Use new sysctl/kvm interfaces. This will stop the "proc size mismatch"
errors when internal kernel structures change size. Also remove
the sgid bit - all live kernel data is accessed through the sysctl
interface.

the PMAP_NEW option is gone as the new interface is no longer optional.

Added experimental fallback /proc-based lookup. Warnings are printed when the
fallback method is used, as the results could be untrustworthy if an
intruder is present. It is highly likely that NetBSD-1.5 will have
an improved kvm interface for reading process information, at which
point this code can be garbage-collected. Also added a word to the
man page -x option description while I was here.

completely remove Mach VM support. all that is left is the all the
header files as UVM still uses (most of) these.

Add CWARNFLAGS+=-Wno-format-y2k.

Kill the CFLAGS+=-Wno-format-y2k which will go, for now, into bsd.sys.mk
It wasn't working anyway, although that could have been fixed by moving it
down below the .include <bsd.prog.mk> line.

It's not completely clear that this usage of %y is in fact y2k-safe, it
seems safer for now to have a single XXX option in bsd.sys.mk, and we need
to rev this anyway for -Wno-uninitialized

This needs strftime()'s %y; use -Wno-format-y2k.

add -DPMAP_NEW if PMAP_NEW is on in mk.conf

add -DUVM to CPPFLAGS if defined(UVM)

Remove WARNS=1 from all the subdirectory Makefiles, and add it to Makefile.inc
now that all /bin has been cleaned.

Fix compiler warnings
Add WARNS=1

LIBMATH -> LIBM (PR# 1688)

Split out some helper functions again.

Create and strvis(3) the argument and environment lists only as needed.
Do not keep extra copies around.

convert to new RCS id conventions.

do not need -I/sys

devname is in libc now

Restore "${DESTDIR}" in CFLAGS.

Need devname.c.

clean up import

needs -lkvm, not -lutil

Fix library dependencies.

Add RCS identifiers.

Don't need -I/sys. (Why was it here?)

changed "Id" to "Header" for rcsids

added rcs ids to all files

branches: 1.1.1;
Initial revision

Clean up deleted files.

clean up import

Add RCS identifiers.

Changed hash_open to dbopen.

changed "Id" to "Header" for rcsids

added rcs ids to all files

branches: 1.1.1;
Initial revision

branches: 1.39.4;
ps: Rename local routine name from uname to usrname

This removes symbol name routine clash with libc's uname(3).
This allows to build ps(1) against LLVM Sanitizers.

Sponsored by <The NetBSD Foundation>

provide a tree like display with -d, from FreeBSD

Calculate CPU usage (pcpu) once per process if it is required. This change
significantly improves performance for slow machines when output is sorted
by pcpu.

ok martin

for donlist{,_sysctl}:
- obtain log_ccpu = log(ccpu) rather than ccpu itself
- use common default values and warn users appropriately when errors occur
ok martin

branches: 1.35.6;
Use an enum type for PRINTMODE vs. WIDTHMODE. Compiler output diffs have
been checked.

branches: 1.34.2;
Add an LTIME column that prints lwp cputime.

branches: 1.33.6; 1.33.12;
Fix ps(1) and top(1) to show reasonable CPU numbers i.e. cpu_index() provided
by the kernel, instead of CPU order number, which is generally random.

Add a CPUID keyword; idea and some of the code from Anon Ymous

add a keyword "lname", to print l_name.

branches: 1.30.8;
Add support for printing the emulation of the process.

* When all columns are given null customised headers, the blank header
line is not printed at all. This is specified in P1003.1-2004
(SUSv3), and is useful.

* Customised headers may contain embedded space, commas and equals
signs. To specify multiple customised headers, use multiple -o or -O
options. This is specified (for "-o", not for "-O") in P1003.1-2004
(SUSv3), and is useful.

* When a column is given a null (blank) customised header, it keeps its
default minimum width. This is specified in P1003.1-2004 (SUSv3), and
seems harmless.

* Fix a bug that made it impossible to print the same keyword multiple
times, with different customised headers each time. (Previously, the
last customised header was used for all instances of the keyword.)

* Make the behaviour of "-O" more useful. The first -O option adds
the default keywords only if there have not yet been any formatting
options, and multiple -O options now insert their keywords in adjacent
positions. Now {ps -j -O %cpu} is like {ps -j} with one extra column;
Previously, it would have had all the columns implied by "-j", plus
all the default columns, plus the extra column specified by "-O".

* Convert from home-grown linked lists to SIMPLEQ lists.

Discussed in tech-userlevel.

sprinkle a little const, and now everything compiles with WARNS=3

Add support for the POSIX.2 "etime" keyword - shows the elapsed time
since the process was started. Fix a couple of style nits as well.

Fix for part of PR standards/11224.

KNF, ANSFify, de-__P, unC++ify, white space nits.

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22249, verified by myself.

- add keywords for utime, stime and ctime (output sss.ssssss if small)
- fiddle with keyword table - mainly to keep width under 80 cols
- add -k keylist so sort output
- don't use logs to work out field width, get -MAXINT right
- user defined headers can be 1 char (actually they can be zero length)
so ?? in tty column isn't guaranteed to be wider than header
- remove some long long casts (compiles on sparc64 still)
- use kvm_openfiles(NULL, NULL, NULL, KVN_NO_FILES, ..) unless core file
specified,
(approved by christos)

Add p_svuid and p_svgid to kinfo_proc2. Populate them in the kernel
and in libkvm. Then teach ps how to show them to you.

Also, teach ps how to show the names for all the uids, the rest of the
group numbers, and the "group access list".

Merge the nathanw_sa branch.

g/c procfs fallback code, it's redundant now that we use sysctl to get
process info

branches: 1.20.2;
Retrieve maxslp & uspace from kernel instead of using defines. Needed for
shared arm code.

Calculate field widths on the fly so that all columns line up nicely
and make more effective use of screen real estate when some columns
(eg USER and VSZ) didn't need the full default width.

G/C evar() - now that everything is in the one structure it was exactly
the same as pvar().

branches: 1.17.2;
Use new sysctl/kvm interfaces. This will stop the "proc size mismatch"
errors when internal kernel structures change size. Also remove
the sgid bit - all live kernel data is accessed through the sysctl
interface.

Since we have an "extern.h", put all external variable declarations in
that file and not sprinkled around various .c files.

"ps -e" only shows the environment for the processes owned by the
current user id or all process if run by root.

Fixes PR security/5967 from Todd Vierling.

When using the procfs for extracting process information, extract
also process start time, process arguments and session leadership status.

The procfs fallback is also used when kvm_openfiles() completely fails
(e.g. when /dev/mem is not readable).

Solves PR 7772, though the final implementation is different.

branches: 1.13.2; 1.13.4;
Added experimental fallback /proc-based lookup. Warnings are printed when the
fallback method is used, as the results could be untrustworthy if an
intruder is present. It is highly likely that NetBSD-1.5 will have
an improved kvm interface for reading process information, at which
point this code can be garbage-collected. Also added a word to the
man page -x option description while I was here.

remove old OLD BSD VM code.

Fix p_nice vs. NZERO code.
Remove some unnecessary casts.
Remove an unused time format field.

Split out some helper functions again.

Create and strvis(3) the argument and environment lists only as needed.
Do not keep extra copies around.

convert to new RCS id conventions.

devname is in libc now

Declare devname().

clean up import

Add RCS identifiers.

changed "Id" to "Header" for rcsids

added rcs ids to all files

branches: 1.1.1;
Initial revision

branches: 1.21.70;
use __RCSID()

branches: 1.20.22;
Wrap some long lines.
Consisently use (void) in front of printf(...);
A few more white-space nits.

KNF, ANSFify, de-__P, unC++ify, white space nits.

branches: 1.18.2;
Just use return value of strvis instead of calling strlen.

Don't define variable as static unnecessarily.

Always initialize an auto variable.

do not update maxlen if realloc fails.

Since we have an "extern.h", put all external variable declarations in
that file and not sprinkled around various .c files.

branches: 1.13.6;
add <sys/time.h> to fix compiler error

RCS Id Police.

Fix compiler warnings
Add WARNS=1

Simplify this just a little.

Split out some helper functions again.

Create and strvis(3) the argument and environment lists only as needed.
Do not keep extra copies around.

Do the right thing for an empty environment.

Avoid an extra (gratuitous) pass through strvis(3).

convert to new RCS id conventions.

Better fix for the previous.

Fix oversight in previous change, when argument list is sort of empty.

kill extra space, save my sanity

branches: 1.1.1;
Initial revision

Don't assign v to newvar, so we can still access the original length.

use emalloc and friends, add ktrace flag printing.

Provide symbolic printing of some keywords by capitalizing them.

Restore maxrss, idrss, isrss, ixrss printing in ps(1)

The RSS related statistics are now back in the NetBSD kernel.

These values were disabled since day0 until today.

libkvm(3) users will still receive inappropriate values as RSS statistics
are updated upon sysctl(3) call.

Patch submitted by <Krzysztof Lasocki>

branches: 1.56.2; 1.56.4;
use EXIT_FAILURE instead of 1

branches: 1.55.2;
ps: Rename local routine name from uname to usrname

This removes symbol name routine clash with libc's uname(3).
This allows to build ps(1) against LLVM Sanitizers.

Sponsored by <The NetBSD Foundation>

Add an LTIME column that prints lwp cputime.

branches: 1.53.6; 1.53.12;
Remove uarea swap-out functionality:

- Addresses the issue described in PR/38828.
- Some simplification in threading and sleepq subsystems.
- Eliminates pmap_collect() and, as a side note, allows pmap optimisations.
- Eliminates XS_CTL_DATA_ONSTACK in scsipi code.
- Avoids few scans on LWP list and thus potentially long holds of proc_lock.
- Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k.
- Removes __SWAP_BROKEN cases.

Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on
acorn26 (thanks to <bjh21>).

Discussed on <tech-kern>, reviewed by <ad>.

Add a CPUID keyword; idea and some of the code from Anon Ymous

add a keyword "lname", to print l_name.

branches: 1.50.8;
Add support for printing the emulation of the process.

c99 initializers

Set the width in the correct data structure. (Fix error in previous.)

* When all columns are given null customised headers, the blank header
line is not printed at all. This is specified in P1003.1-2004
(SUSv3), and is useful.

* Customised headers may contain embedded space, commas and equals
signs. To specify multiple customised headers, use multiple -o or -O
options. This is specified (for "-o", not for "-O") in P1003.1-2004
(SUSv3), and is useful.

* When a column is given a null (blank) customised header, it keeps its
default minimum width. This is specified in P1003.1-2004 (SUSv3), and
seems harmless.

* Fix a bug that made it impossible to print the same keyword multiple
times, with different customised headers each time. (Previously, the
last customised header was used for all instances of the keyword.)

* Make the behaviour of "-O" more useful. The first -O option adds
the default keywords only if there have not yet been any formatting
options, and multiple -O options now insert their keywords in adjacent
positions. Now {ps -j -O %cpu} is like {ps -j} with one extra column;
Previously, it would have had all the columns implied by "-j", plus
all the default columns, plus the extra column specified by "-O".

* Convert from home-grown linked lists to SIMPLEQ lists.

Discussed in tech-userlevel.

Coverity CID 1559: Fix memory leak.

add a keyword for the uarea address. very useful with crashdumps.

sprinkle a little const, and now everything compiles with WARNS=3

branches: 1.43.2;
print TTY instead of TT, one question mark instead of two and always the
full tty name.

Add a "laddr" keyword, analagous to "paddr". Useful for finding
struct pcb's.

For the "lid" keyword, change ULONG to INT32 since it's a 32bit type.
From Nicolas Joly in PR bin/26452.

branches: 1.40.2;
Add support for the POSIX.2 "args" and "comm" keywords. "args" is just
an alias for "command", and "comm" shows just argv[0] and not the whole
argument list.

Fix for part of PR standards/11224.

Add support for the POSIX.2 "etime" keyword - shows the elapsed time
since the process was started. Fix a couple of style nits as well.

Fix for part of PR standards/11224.

Wrap some long lines.
Consisently use (void) in front of printf(...);
A few more white-space nits.

Change the header for the "tty" keyword "TT" instead of "TTY".

Fixed POSIX.2 conformance problem mentioned in standards/11225 from
Ben Harris.

Correct typo in page header for tpgid.

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22249, verified by myself.

Support -ktt and -ktty, quote "tdev" in comment so table can be sorted.

XXX: cast tdev to INT32 instead of UINT32 so that that sorting works as
before [NODEV entries come first]

- add keywords for utime, stime and ctime (output sss.ssssss if small)
- fiddle with keyword table - mainly to keep width under 80 cols
- add -k keylist so sort output
- don't use logs to work out field width, get -MAXINT right
- user defined headers can be 1 char (actually they can be zero length)
so ?? in tty column isn't guaranteed to be wider than header
- remove some long long casts (compiles on sparc64 still)
- use kvm_openfiles(NULL, NULL, NULL, KVN_NO_FILES, ..) unless core file
specified,
(approved by christos)

Seems better to display p_sv{u,g}id when asked for it...

Add p_svuid and p_svgid to kinfo_proc2. Populate them in the kernel
and in libkvm. Then teach ps how to show them to you.

Also, teach ps how to show the names for all the uids, the rest of the
group numbers, and the "group access list".

Merge the nathanw_sa branch.

correct use of format string (signedness, %ldd). PR 18679

branches: 1.27.2;
Since kinfo_proc2 has many values that are 64-bit, ps needs to know
that they're 64-bit, and grab them out of memory appropriately. Otherwise,
big-endian systems get the wrong end of the 64-bit value and lose.

Keywords affected: inblk, majflt, minflt, msgrcv, msgsnd, nivcsw, nsigs,
nswap, nvcsw, and oublk.

branches: 1.26.2;
address PR 11653. checked on alpha.
XXX supports 32bit and 64bit pointers only

branches: 1.25.2;
Calculate field widths on the fly so that all columns line up nicely
and make more effective use of screen real estate when some columns
(eg USER and VSZ) didn't need the full default width.

G/C evar() - now that everything is in the one structure it was exactly
the same as pvar().

branches: 1.23.2;
Use new sysctl/kvm interfaces. This will stop the "proc size mismatch"
errors when internal kernel structures change size. Also remove
the sgid bit - all live kernel data is accessed through the sysctl
interface.

Fix problem in option processing, would segfault when parsing "-O,".
Closes PR/10096 by Yuji Yamano using the patch provided.

Since we have an "extern.h", put all external variable declarations in
that file and not sprinkled around various .c files.

Since our gcc doesn't warn about NULL format strings anymore, we can
fix the incorrect err(1, "%s", "") et al.
Closes PR bin/7592 by cgd.

branches: 1.19.4;
limit session to 6 chars. from Geoff Wing <gcw@pobox.com> and fixes his PR#7506

Add support for printing wide sigset_t properly.

Add a "sid" keyword to print the session ID.

remove old OLD BSD VM code.

Fix compiler warnings
Add WARNS=1

Fix p_nice vs. NZERO code.
Remove some unnecessary casts.
Remove an unused time format field.

in findvar(), return after warning about unknown keyword so we don't
deref. a null pointer.

branches: 1.12.6;
correct the sizes specified for various variables.

Kenneth Stailey pointed out that his previous fix was incorrect. RE and SL
should be bounded at 127, not 99.

as suggested in PR 1347, by Kenneth Stailey, make "RE" and "SL" fields
show up as two digits, max, displaying "99" if > 99. allow any field with
an unsigned value to be specified as printing that way (though right now
it's only used for RE and SL, which are both unsigned longs). This fix
is substantially different than that suggested in the PR.

add a 'holdcnt' keyword and document it.

convert to new RCS id conventions.

clean up import

struct element foo

update for changed flags

Add RCS identifiers.

changed "Id" to "Header" for rcsids

added rcs ids to all files

branches: 1.1.1;
Initial revision

Don't rely on the USPACE kernel define as a fallback if the vm.uspace
sysctl fails. We've got bigger problems if the sysctl fail anyway.

branches: 1.27.14;
for donlist{,_sysctl}:
- obtain log_ccpu = log(ccpu) rather than ccpu itself
- use common default values and warn users appropriately when errors occur
ok martin

branches: 1.26.12; 1.26.46;
Remove clause 3 and 4 from TNF licenses

branches: 1.25.16;
c99 initializers

Use HW_PHYSMEM64 to avoid problem with > 2GB memory (and >4GB as well)
Fixes PR/23717 and PR/24092

#define<TAB>

KNF, ANSFify, de-__P, unC++ify, white space nits.

donlist_sysctl() was overwriting maxslp with vm.uspace. Changed
it to use uspace. See PR bin/23713, approved by christos@netbsd.org.
Thanks to joff@stchome.com for the submission.

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22249, verified by myself.

Merge the nathanw_sa branch.

branches: 1.18.2;
Retrieve maxslp & uspace from kernel instead of using defines. Needed for
shared arm code.

Fix handling of "ps U<user>" where <user> ended it 't' - now if the
first argument doesn't start with a '-' and ends in 't', the 't' only
gets converted to a 'T' (for listing processes on the current tty)
if it doesn't contain any other command line option characters that
take an argument. Problem noticed by ITOH Yasufumi on tech-userlevel.

If the 'start' format was the last column, the width wouldn't be set
and started() output a zero length string instead of a string the right
length - fix header width calculations so the the last column is handled
the same all others. This only showed up for 'start' because started()
knew that the column header is the same width and the field itself...

Update copyrights.

branches: 1.16.2;
Use new sysctl/kvm interfaces. This will stop the "proc size mismatch"
errors when internal kernel structures change size. Also remove
the sgid bit - all live kernel data is accessed through the sysctl
interface.

Since we have an "extern.h", put all external variable declarations in
that file and not sprinkled around various .c files.

branches: 1.14.6;
Use the physmem variable to get at the physical memory size (instead of the
avail_end - avail_start difference which is only correct on machines with
contiguous memory).

remove old OLD BSD VM code.

Fix compiler warnings
Add WARNS=1

convert to new RCS id conventions.

from James Jegers <jimj@miller.cs.uwm.edu>: quiet -Wall, and squelch
some of the worst style errors.

Format police.

print %MEM correctly

clean up import

update for changed flags

Add RCS identifiers.

changed "Id" to "Header" for rcsids

added rcs ids to all files

after 0.2.2 "stable" patches applied

branches: 1.1.1;
Initial revision

remove double t from targeted, add missing r to arbitrary
And fix few more typos along the way in comments and man pages.

use emalloc and friends, add ktrace flag printing.

Provide symbolic printing of some keywords by capitalizing them.

Remove SCCS workaround. No binary change.

- dedup code
- add a safe_strftime() to handle error cases

Fix the column width calculation for the lstart column if an empty
column header is specified.

Fixes bug pointed out by Ted Spradley in
https://mail-index.netbsd.org/netbsd-users/2021/04/05/msg026808.html .

branches: 1.132.2;
Add a fallback definition of LSDEAD in ps(1)

The symbol is no longer available in headers.

Requested by <mrg>

Make LSDEAD usage conditional

LSDEAD is not used since NetBSD-5.0 and will be gone.

The same conditional usage is already in ps.c in the same program.

Don't display l_wchan, either there is something in l_wmesg and we display
it, or there's nothing and we print "-".

branches: 1.129.2;
use EXIT_FAILURE instead of 1

branches: 1.128.2;
ps: Rename local routine name from uname to usrname

This removes symbol name routine clash with libc's uname(3).
This allows to build ps(1) against LLVM Sanitizers.

Sponsored by <The NetBSD Foundation>

branches: 1.127.6;
Handle functions that use the offset from either kinfo_proc2/kinfo_lwp properly.

provide a tree like display with -d, from FreeBSD

Calculate CPU usage (pcpu) once per process if it is required. This change
significantly improves performance for slow machines when output is sorted
by pcpu.

ok martin

for donlist{,_sysctl}:
- obtain log_ccpu = log(ccpu) rather than ccpu itself
- use common default values and warn users appropriately when errors occur
ok martin

branches: 1.123.2;
Use l_wmesg if the string is not empty. Don't bother checking l_name for
nullness.

branches: 1.122.2;
Use an enum type for PRINTMODE vs. WIDTHMODE. Compiler output diffs have
been checked.

branches: 1.121.2;
Add an LTIME column that prints lwp cputime.

branches: 1.120.2;
Use C89 function definitions

Remove unused variable.
From cppcheck via Henning Petersen in PR 46002.

branches: 1.118.2;
When converting from pages to kilobytes, cast the return value of
getpagesize() to size_t. For some reason getpagesize() is defined to
return int, and several of the page counts we get come back from the
kernel as int32_t; in LP64 without the cast the byte count will be
computed in a 32-bit value and for large processes will overflow and
become negative... and then remain negative when divided by 1024 to
convert to kilobytes.

Fixes a problem I hit the other day where I saw negative RSS, which
turns out also to be PR 40642.

Note: other logic in here will break down when we first get >2TB
processes... and int32 page counts will break on >8TB processes. But
hopefully we won't see any of that for a few years yet.

branches: 1.117.2;
fix previous.

Make printing of lwp flags similar to the process one, identifying (O)nproc
(K)ernel threads and scheduler (a)ctivations.

branches: 1.115.2;
Do not check pcpu value against 100.0 but rather 99.95, to properly
deal with value rounding. From apb@.

Make %cpu drop the decimal part when reaching 100%, to stay in the 5
expected columns.

Fix ps(1) and top(1) to show reasonable CPU numbers i.e. cpu_index() provided
by the kernel, instead of CPU order number, which is generally random.

Remove uarea swap-out functionality:

- Addresses the issue described in PR/38828.
- Some simplification in threading and sleepq subsystems.
- Eliminates pmap_collect() and, as a side note, allows pmap optimisations.
- Eliminates XS_CTL_DATA_ONSTACK in scsipi code.
- Avoids few scans on LWP list and thus potentially long holds of proc_lock.
- Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k.
- Removes __SWAP_BROKEN cases.

Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on
acorn26 (thanks to <bjh21>).

Discussed on <tech-kern>, reviewed by <ad>.

- add new RLIMIT_AS (aka RLIMIT_VMEM) resource that limits the total
address space available to processes. this limit exists in most other
modern unix variants, and like most of them, our defaults are unlimited.
remove the old mmap / rlimit.datasize hack.

- adds the VMCMD_STACK flag to all the stack-creation vmcmd callers.
it is currently unused, but was added a few years ago.

- add a pair of new process size values to kinfo_proc2{}. one is the
total size of the process memory map, and the other is the total size
adjusted for unused stack space (since most processes have a lot of
this...)

- patch sh, and csh to notice RLIMIT_AS. (in some cases, the alias
RLIMIT_VMEM was already present and used if availble.)

- patch ps, top and systat to notice the new k_vm_vsize member of
kinfo_proc2{}.

- update irix, svr4, svr4_32, linux and osf1 emulations to support
this information. (freebsd could be done, but that it's best left
as part of the full-update of compat/freebsd.)


this addresses PR 7897. it also gives correct memory usage values,
which have never been entirely correct (since mmap), and have been
very incorrect since jemalloc() was enabled.

tested on i386 and sparc64, build tested on several other platforms.

thanks to many folks for feedback and testing but most espcially
chuq and yamt for critical suggestions that lead to this patch not
having a special ugliness i wasn't happy with anyway :-)

fix -Wsign-compare issues

branches: 1.109.2;
make sure the table stays aligned even if no groups are to be displayed
with the "groupnames" option

make sure the table stays aligned even if no groups are to be displayed
with the "groups" option

fix dev_t 64 bit issues.

branches: 1.106.2;
Use 'O' for LSONPROC like solaris instead of bundling LSIDL, LSRUN, LSONPROC
to 'R'

Remove clause 3 and 4 from TNF licenses

branches: 1.104.4;
Add a CPUID keyword; idea and some of the code from Anon Ymous

Remove systrace. Ok core@.

Kill another use of PZERO.

add a keyword "lname", to print l_name.

branches: 1.100.4; 1.100.6;
Don't call P_ZOMBIE(struct kinfo_proc2), it's p_stat field contains
lwp states, not process states!
Since the times are filled with zeros for zombie processes the check
before assuming teh process time is zero is moot - delete it.
Fix the other chact to test the p_realstat.
Fixer PR kern/36295

Revert the changes done after newlock2 merge, except for the removal of
the now unused LSDEAD state. The current kernel headers are now
source-compatible with pre-newlock2 kernel headers again.

Approved by Andrew Doran.

Further changes to use compat process flags mapped by sysctl. To be
revisited soon.

Sync with kernel changes introduced by merging the newlock2 branch.

branches: 1.96.2; 1.96.4;
Add support for printing the emulation of the process.

* When all columns are given null customised headers, the blank header
line is not printed at all. This is specified in P1003.1-2004
(SUSv3), and is useful.

* Customised headers may contain embedded space, commas and equals
signs. To specify multiple customised headers, use multiple -o or -O
options. This is specified (for "-o", not for "-O") in P1003.1-2004
(SUSv3), and is useful.

* When a column is given a null (blank) customised header, it keeps its
default minimum width. This is specified in P1003.1-2004 (SUSv3), and
seems harmless.

* Fix a bug that made it impossible to print the same keyword multiple
times, with different customised headers each time. (Previously, the
last customised header was used for all instances of the keyword.)

* Make the behaviour of "-O" more useful. The first -O option adds
the default keywords only if there have not yet been any formatting
options, and multiple -O options now insert their keywords in adjacent
positions. Now {ps -j -O %cpu} is like {ps -j} with one extra column;
Previously, it would have had all the columns implied by "-j", plus
all the default columns, plus the extra column specified by "-O".

* Convert from home-grown linked lists to SIMPLEQ lists.

Discussed in tech-userlevel.

Add some more const so that this compiles again.

sprinkle a little const, and now everything compiles with WARNS=3

Appease -Wuninitialized by initializing val and uval to 0.
Marked with XXXGCC for hpcarm (found while building for it).

Reviewed (in a slightly different form) by lukem.

print TTY instead of TT, one question mark instead of two and always the
full tty name.

s/the the/the/ (only in sources that aren't regularly imported from
elsewhere).

Gah, fix uninitialised variable with the "etime" change. _Never_ be
happy with something and then edit a single line before committing
without retesting...

Add support for the POSIX.2 "args" and "comm" keywords. "args" is just
an alias for "command", and "comm" shows just argv[0] and not the whole
argument list.

Fix for part of PR standards/11224.

Add support for the POSIX.2 "etime" keyword - shows the elapsed time
since the process was started. Fix a couple of style nits as well.

Fix for part of PR standards/11224.

#define<TAB>

Use some constants from <tzfile.h>.

Wrap some long lines.
Consisently use (void) in front of printf(...);
A few more white-space nits.

Fix grammar in a comment.

KNF, ANSFify, de-__P, unC++ify, white space nits.

Spell interrupt with two rs. From Peter Postma.

g/c SDEAD

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22249, verified by myself.

- add keywords for utime, stime and ctime (output sss.ssssss if small)
- fiddle with keyword table - mainly to keep width under 80 cols
- add -k keylist so sort output
- don't use logs to work out field width, get -MAXINT right
- user defined headers can be 1 char (actually they can be zero length)
so ?? in tty column isn't guaranteed to be wider than header
- remove some long long casts (compiles on sparc64 still)
- use kvm_openfiles(NULL, NULL, NULL, KVN_NO_FILES, ..) unless core file
specified,
(approved by christos)

Add p_svuid and p_svgid to kinfo_proc2. Populate them in the kernel
and in libkvm. Then teach ps how to show them to you.

Also, teach ps how to show the names for all the uids, the rest of the
group numbers, and the "group access list".

Merge the nathanw_sa branch.

interrupt with two rs.

g/c procfs fallback code, it's redundant now that we use sysctl to get
process info

Document some missing flags and add systrace(1) 'x' flag.

Since kinfo_proc2 has many values that are 64-bit, ps needs to know
that they're 64-bit, and grab them out of memory appropriately. Otherwise,
big-endian systems get the wrong end of the 64-bit value and lose.

Keywords affected: inblk, majflt, minflt, msgrcv, msgsnd, nivcsw, nsigs,
nswap, nvcsw, and oublk.

branches: 1.71.2;
Use int32_t for temporary variables (instead of long), that is the same
type as we get the data passed from the kernel. This avoids (missing)
sign extension bugs on LP64 systems and partly takes care of PR 15677.

We now print this values as negative seconds - still wrong, but that
probably is due to the simple way this values are acumulated in the
scheduler, causing negative times when ntpd steps time backwards.

started(): don't assume that the field width is correct and recompute
it as needed

This fixes bogus output for cases like `ps -ao start=xx`, as reported by
Martin Macok of Undeground.cz on current-users@netbsd.org.

fix comment: "[]" are [square] brackets, "<>" are angle brackets

Print commands within angled brackets only if they are system processes.
Print other commands, like cc, cpp and as, within parentheres.

Print system processes/kernel threads as [kprocname] as top(1) does.

Retrieve maxslp & uspace from kernel instead of using defines. Needed for
shared arm code.

clarify things a bit more (from hubertf)

Really explain what parenthesized names mean and why we parenthesize names!

Document the meaning of commands shown in parantheses.

Patches submitted by Jeremy C. Reed <reed@reedmedia.net> in PR 11930.

KPTR/KPTR24 are using u_int64_t. commented by minoura

redo previous (simplify KPTR)

backout previous; did not compile on non-alpha

simplify KPTR processing.

address PR 11653. checked on alpha.
XXX supports 32bit and 64bit pointers only

Add a K flag to the state information - K is for kernel thread or system
process (ie, P_SYSTEM set in p_flag). The L flag (locked in core) is
only used for non-P_SYSTEM processes.

remove include of <vm/vm.h>

branches: 1.55.2;
Fix handling of "ps U<user>" where <user> ended it 't' - now if the
first argument doesn't start with a '-' and ends in 't', the 't' only
gets converted to a 'T' (for listing processes on the current tty)
if it doesn't contain any other command line option characters that
take an argument. Problem noticed by ITOH Yasufumi on tech-userlevel.

If the 'start' format was the last column, the width wouldn't be set
and started() output a zero length string instead of a string the right
length - fix header width calculations so the the last column is handled
the same all others. This only showed up for 'start' because started()
knew that the column header is the same width and the field itself...

Update copyrights.

Fix back-to-front assignment in tname() which meant that the column
width would never autosize past the header ("TT") width. Problem
noticed by Geoff Wing.

Calculate field widths on the fly so that all columns line up nicely
and make more effective use of screen real estate when some columns
(eg USER and VSZ) didn't need the full default width.

G/C evar() - now that everything is in the one structure it was exactly
the same as pvar().

branches: 1.51.2;
Use new sysctl/kvm interfaces. This will stop the "proc size mismatch"
errors when internal kernel structures change size. Also remove
the sgid bit - all live kernel data is accessed through the sysctl
interface.

Handle SONPROC state.

Fix CL* purge fallout.

Since we have an "extern.h", put all external variable declarations in
that file and not sprinkled around various .c files.

"ps -e" only shows the environment for the processes owned by the
current user id or all process if run by root.

Fixes PR security/5967 from Todd Vierling.

command(): don't start using procfs_getargv() instead of kvm_getargv()
just because the latter returned a NULL - that may normally happen,
when the process is, for example, a kernel thread

procfs_getargv(): if the name is brackened in '(' and ')', just
return NULL immediately - the code in command() will DTRT
and the output will be same for procfs- and kvm- based lookup
in all cases

When using the procfs for extracting process information, extract
also process start time, process arguments and session leadership status.

The procfs fallback is also used when kvm_openfiles() completely fails
(e.g. when /dev/mem is not readable).

Solves PR 7772, though the final implementation is different.

skip "dty" just like we skip "tty", otherwise we just get "dt" in ps output for all dty's

branches: 1.43.2;
According to the previous commit the new process state should be SDEAD and
not SDYING, so s/SDYING/SDEAD/.

Update for SDEAD.

limit session to 6 chars. from Geoff Wing <gcw@pobox.com> and fixes his PR#7506

Add support for printing wide sigset_t properly.

branches: 1.39.2;
Look for the formats used by setproctitle(3) and login shells, and avoiding
outputting the redundant `(comm)'.

__AUDIT__ cleanup.

Need <time.h> for localtime(), time() and strftime() prototypes.

remove old OLD BSD VM code.

Begone KERNBASE! (Just print kernel pointers as-is.)

use %c not %C for lstart; from tetsuya@secom-sis.co.jp via OpenBSD

Fix compiler warnings
Add WARNS=1

DTRT if e_wmesg is not NUL-terminated.

DTRT if e_login is not NUL-terminated.

Fix p_nice vs. NZERO code.
Remove some unnecessary casts.
Remove an unused time format field.

return 0.0 cpu usage for zombies. closes pr bin/2693. Perhaps this should
be fixed in exit, so p->p_pctcpu is cleared when the process exits.

include <pwd.h> for one function (user_from_uid) and correct one
%*d -> %*ld in a format string. (yes, trying to find the problem with
ps wwwwwwwww1 and core dump, and might as well clean up while here)

add code to convert from INT32 and UINT32 to native types.

Kenneth Stailey pointed out that his previous fix was incorrect. RE and SL
should be bounded at 127, not 99.

as suggested in PR 1347, by Kenneth Stailey, make "RE" and "SL" fields
show up as two digits, max, displaying "99" if > 99. allow any field with
an unsigned value to be specified as printing that way (though right now
it's only used for RE and SL, which are both unsigned longs). This fix
is substantially different than that suggested in the PR.

typeof(timeval.ts_sec) != time_t

If there is another field after the command, fill in the command field with
blanks.

Split out some helper functions again.

Create and strvis(3) the argument and environment lists only as needed.
Do not keep extra copies around.

Avoid an extra (gratuitous) pass through strvis(3).

add the ability to understand and print INT and UINT-sized variables.

convert to new RCS id conventions.

be careful with addresses.

from James Jegers <jimj@miller.cs.uwm.edu>: quiet -Wall, and squelch
some of the worst style errors.

kill UPAGES

NBPG -> getpagesize()

No more P_NOSWAP/P_PHYSIO.

clean up import

struct element foo

update for changed flags

Fix test for zombie process.

Add RCS identifiers.

use strvis on the arguments and environment, so we don't end up in
sed and/or xterm-title-changing hell.

make ps do something more sensible with zombie processes

fix more-or-less-uninitialized variable...

add support and documentation for the 'e' option, to print out
the process environment.

changed "Id" to "Header" for rcsids

added rcs ids to all files

branches: 1.1.1;
Initial revision

g/c procfs fallback code, it's redundant now that we use sysctl to get
process info

branches: 1.9.6;
Use new sysctl/kvm interfaces. This will stop the "proc size mismatch"
errors when internal kernel structures change size. Also remove
the sgid bit - all live kernel data is accessed through the sysctl
interface.

Ignore also "self" link when searching for proc info.
Patch sent by NAKAJIMA Yoshihiro in bin/9819

command(): don't start using procfs_getargv() instead of kvm_getargv()
just because the latter returned a NULL - that may normally happen,
when the process is, for example, a kernel thread

procfs_getargv(): if the name is brackened in '(' and ')', just
return NULL immediately - the code in command() will DTRT
and the output will be same for procfs- and kvm- based lookup
in all cases

When using the procfs for extracting process information, extract
also process start time, process arguments and session leadership status.

The procfs fallback is also used when kvm_openfiles() completely fails
(e.g. when /dev/mem is not readable).

Solves PR 7772, though the final implementation is different.

branches: 1.5.2;
KNF.

branches: 1.4.2;
- Be really paranoid. Any time we open a file that is in /proc, run
fstatfs() on the open file to verify that no intruder has mounted
something on portions of /proc. This will catch, for instance,
"mount_null /tmp /proc/1378". We already do 5 syscalls per
process, so one more won't hurt :) ... and safety is better than
performance when ps is otherwise broken.
- Also added a few close()'s at early returns, to avoid chewing up fd's.

Minor fixes:
1. Use statfs() to verify that /proc is a procfs.
2. Add -K option to disable /proc-based method.
3. Make warnings less verbose, but still accurate.

Add missing prototype so that this actually compiles.

Added experimental fallback /proc-based lookup. Warnings are printed when the
fallback method is used, as the results could be untrustworthy if an
intruder is present. It is highly likely that NetBSD-1.5 will have
an improved kvm interface for reading process information, at which
point this code can be garbage-collected. Also added a word to the
man page -x option description while I was here.

s/maxiumum/maximum/

reflect reality on flags.

Provide symbolic printing of some keywords by capitalizing them.

Sort options in synopsis.

PR standards/11223

Add -G to take a single group argument heading towards POSIX.2 compliance.
Patch from jperkin and reviewed by simonb.

branches: 1.110.6;
Restore maxrss, idrss, isrss, ixrss printing in ps(1)

The RSS related statistics are now back in the NetBSD kernel.

These values were disabled since day0 until today.

libkvm(3) users will still receive inappropriate values as RSS statistics
are updated upon sysctl(3) call.

Patch submitted by <Krzysztof Lasocki>

branches: 1.109.4; 1.109.6;
Bump date for previous. Remove/replace bogus Tn uses.

Remove the filesystem tracing feature

This is a legacy interface from 4.4BSD, and it was
introduced to overcome shortcomings of ptrace(2) at that time, which are
no longer relevant (performance). Today /proc/#/ctl offers a narrow
subset of ptrace(2) commands and is not applicable for modern
applications use beyond simplistic tracing scenarios.

This removal will simplify kernel internals. Users will still be able to
use all the other /proc files.

This change won't affect other procfs files neither Linux compat
features within mount_procfs(8). /proc/#/ctl isn't available on Linux.

Remove:
- /proc/#/ctl from mount_procfs(8)
- P_FSTRACE note from the documentation of ps(1)
- /proc/#/ctl and filesystem tracing documentation from mount_procfs(8)
- KAUTH_REQ_PROCESS_PROCFS_CTL documentation from kauth(9)
- source code file miscfs/procfs/procfs_ctl.c
- PFSctl and procfs_doctl() from sys/miscfs/procfs/procfs.h
- KAUTH_REQ_PROCESS_PROCFS_CTL from sys/sys/kauth.h
- PSL_FSTRACE (0x00010000) from sys/sys/proc.h
- P_FSTRACE (0x00010000) from sys/sys/sysctl.h

Reduce code complexity after removal of this functionality.

Update TODO.ptrace accordingly: remove two entries about /proc tracing.

Do not keep legacy notes as comments in the headers about removed
PSL_FSTRACE / P_FSTRACE, as this interface had little number of users
(close or equal to zero).

Proposed on tech-kern@.

All filesystem tracing utility users are encouraged to switch to ptrace(2).

Sponsored by <The NetBSD Foundation>

Remove workaround for ancient HTML generation code.

branches: 1.106.6;
provide a tree like display with -d, from FreeBSD

Earliest reference in the TUHS archive is v3
Unfortunately only the source code for the compiler is present for that
release. The v2 archive is missing the manuals & only contains a some of the
commands.
Switch to v3 and reference manual category eight, as per r1.32 of
src/bin/ps/ps.1 from OpenBSD.
Heads up from Ingo Schwarze.

Document the version ps first appeared.

branches: 1.103.8;
Bump date for previous.

Add an LTIME column that prints lwp cputime.

branches: 1.101.2;
Switch device database to cdb(5). Rework ttyname(3) and ttyname_r(3) to
depend on new devname_r(3) as heart. Add /dev/pts magic directly to
devname(3). While it can lead to returning non-existing paths, the
behavior is more consistent that way. Drop caching layer in devname(3),
it doesn't buy anything for the common case of having access to the
database. Teach devname(3) proper fallback behavior of scanning /dev.
Create both old-style and new-style database for now in /etc/rc.d/sysdb.

Increase width of table column so it formats more nicely.

- choose better argument name for `-U';
- add argument names for options in man page;
- improve wording, punctuation, capitalization.

From Bug Hunting in email.

Remove trailing whitespace.

Fix obsolete example (that referred to letters in process state that are
not set anymore), and clarify their meaning.

The 'lstart' column uses strftime %c, not %C. Noted by mrg.

branches: 1.95.2;
Minor English improvements for -w, partly from Snader_LB.

branches: 1.94.4;
fix typo, from Randolf Richardson in PR 44559

branches: 1.93.2;
Don't break lines in .It, there be dragons in groff.

Use .In instead of .Aq Pa for header files

Bump date for uarea swap-out removal.

Remove uarea swap-out functionality:

- Addresses the issue described in PR/38828.
- Some simplification in threading and sleepq subsystems.
- Eliminates pmap_collect() and, as a side note, allows pmap optimisations.
- Eliminates XS_CTL_DATA_ONSTACK in scsipi code.
- Avoids few scans on LWP list and thus potentially long holds of proc_lock.
- Cuts ~1.5k lines of code. Reduces amd64 kernel size by ~4k.
- Removes __SWAP_BROKEN cases.

Tested on x86, mips, acorn32 (thanks <mpumford>) and partly tested on
acorn26 (thanks to <bjh21>).

Discussed on <tech-kern>, reviewed by <ad>.

Bump date for previous.

Remove reference to P_SYSTRACE.
Fixes PR/41426
(leave defn in sys/sysctl.h since the bit shouldn't be reused!)

Fix number of columns in .Bl. Escape %C.

branches: 1.86.4;
Improve wording.

Use 'O' for LSONPROC like solaris instead of bundling LSIDL, LSRUN, LSONPROC
to 'R'

Sort options.
Remove systrace reference.

Add "ps -A" option. This displays information about all processes,
like BSD "ps -ax" or SYSV "ps -ef". It's specified in SUSv3.

Also make some minor clarifications to the man page.

Add a CPUID keyword; idea and some of the code from Anon Ymous

Remove systrace. Ok core@.

Document "emul" keyword. Fixes PR 36935.
Also sort list of keywords, and bump date.

add a keyword "lname", to print l_name.

branches: 1.78.4;
remove descriptions of some states. ('A', 'S', '>')
- they are wrong. (madvise is not vadvise)
- they don't exist anymore.

Remove trailing whitespace.
Remove a partially duplicated sentence.

Make it clear that multiple sort keys may be specified.
Closes PR 34271.

Avoid .No if easily possible.

Add some markup missed in previous.

Document changes to -o and -O options, and improve markup.

add "uaddr" and improve text for "laddr".

branches: 1.71.2;
Add the missing suspended state 'U'.

Bump date.

Bump date for laddr.

Add a "laddr" keyword, analagous to "paddr". Useful for finding
struct pcb's.

s/the the/the/ (only in sources that aren't regularly imported from
elsewhere).

Add support for the POSIX.2 "args" and "comm" keywords. "args" is just
an alias for "command", and "comm" shows just argv[0] and not the whole
argument list.

Fix for part of PR standards/11224.

Add support for the POSIX.2 "etime" keyword - shows the elapsed time
since the process was started. Fix a couple of style nits as well.

Fix for part of PR standards/11224.

Uppercase CPU where appropriate.

Add Xr for dev_mkdb(8), per PR 23168. Increase width in FILES section while here.

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22249, verified by myself.

More mdoc.

Update date for recent changes.

- add keywords for utime, stime and ctime (output sss.ssssss if small)
- fiddle with keyword table - mainly to keep width under 80 cols
- add -k keylist so sort output
- don't use logs to work out field width, get -MAXINT right
- user defined headers can be 1 char (actually they can be zero length)
so ?? in tty column isn't guaranteed to be wider than header
- remove some long long casts (compiles on sparc64 still)
- use kvm_openfiles(NULL, NULL, NULL, KVN_NO_FILES, ..) unless core file
specified,
(approved by christos)

Add p_svuid and p_svgid to kinfo_proc2. Populate them in the kernel
and in libkvm. Then teach ps how to show them to you.

Also, teach ps how to show the names for all the uids, the rest of the
group numbers, and the "group access list".

.Nm does not need a dummy argument ("") before punctuation or
for correct formatting of the SYNOPSIS any longer.

Merge the nathanw_sa branch.

Use more mdoc.

New policy: New sentences start on a new line.
Patches by Robert Elz <kre at munnari oz au>, with minimal changes by me.

g/c procfs fallback code, it's redundant now that we use sysctl to get
process info

Sort SEE ALSO.

Document some missing flags and add systrace(1) 'x' flag.

"a, b, c, and d", not "a, b, c and d".

branches: 1.50.2;
Drop trailing space.

Xref pkill/pgrep.

Generate <>& symbolically. I'm avoiding .../dist/... directories for now.

Properly quote punctuation.

Whitespace nits.

precede, not preceed.

Print system processes/kernel threads as [kprocname] as top(1) does.

correct the description of paddr

clarify things a bit more (from hubertf)

Really explain what parenthesized names mean and why we parenthesize names!

Document the meaning of commands shown in parantheses.

Patches submitted by Jeremy C. Reed <reed@reedmedia.net> in PR 11930.

Add a K flag to the state information - K is for kernel thread or system
process (ie, P_SYSTEM set in p_flag). The L flag (locked in core) is
only used for non-P_SYSTEM processes.

Update description of process flag 0x20000 to match reality:
P_NOCLDWAIT, not P_SSTEP.

Remove the 'poip' (pageouts in progress) keyword from the manpage.
From PR bin/8843 by Brian Stark.

branches: 1.36.4;
Use tty "?" to select processes with no controlling tty and tty "-" to
select processes with a revoked controlling tty.

Idea from SunOS/Solaris, suggested by Alexis Rosen.

make -M imply -K. the current system is not likely to be
a reasonable approximation of a crash dump.

name database is kvm.db not kvm_netbsd.db

Document -U switch (displays processes belonging to given user/uid).
Reported in PR 8943 by WL <wliao@dura.spc.uchicago.edu>

"ps -e" only shows the environment for the processes owned by the
current user id or all process if run by root.

Fixes PR security/5967 from Todd Vierling.

branches: 1.31.6;
Minor fixes:
1. Use statfs() to verify that /proc is a procfs.
2. Add -K option to disable /proc-based method.
3. Make warnings less verbose, but still accurate.

Added experimental fallback /proc-based lookup. Warnings are printed when the
fallback method is used, as the results could be untrustworthy if an
intruder is present. It is highly likely that NetBSD-1.5 will have
an improved kvm interface for reading process information, at which
point this code can be garbage-collected. Also added a word to the
man page -x option description while I was here.

More and more .Os cleanups. .Os is defined in the tmac.doc-common file,
so we shouldn't override it with versions in the manpages. Many more to
come.

ps's information is sorted by controlling terminal and (among processes with
the same controlling terminal) by PID.

Clean up SYNOPSIS formatting.

Fix formatting errors.

The system namelist database is called "/var/db/kvm.db" not
"/var/db/kvm_netbsd.db". Fixes PR bin/6712 by Brian Grayson.

change references to vadvise to madvise; remove reference to lisp(1)

Sync with reality: the uprocp keyword doesn't exist.

remove old OLD BSD VM code.

branches: 1.21.2;
Use .Nm "" instead of .Nm ps in SYNOPSIS.

all but the first .Nm in SYNOPSIS need to be spelt out. [bin/4256]

getopt returns -1 not EOF
cleanup manpage

default swap device is /dev/drum, not nonexistent /dev/swap; from
<entropy@tardis.bernstein.com> in PR bin/4042

Refer to execve(2), add section to `vfork' ref. (almost per PR#3627).

fix damage caused by sccs (long ago)

add a 'holdcnt' keyword and document it.

Sync process flags with sys/proc.h

convert to new RCS id conventions.

Integrate `-c' option, from Arne Juul.

vmunix -> netbsd (PR #375)

branches: 1.10.2;
clean up import

Fix spelling errors.

fix from Thomas Eberhardt <thomas@mathematik.uni-Bremen.de>:
don't try to display the resident text size; we can't do that anymore

Add RCS indentifiers.

Fix up FILES sections.

Update file names.

add support and documentation for the 'e' option, to print out
the process environment.

changed "Id" to "Header" for rcsids

added rcs ids to all files

branches: 1.1.1;
Initial revision

ps: remove unreachable break statement after return statement

No binary change.

branches: 1.97.4;
use emalloc and friends, add ktrace flag printing.

use parsenum like everywhere else in the code.

Sort options in usage.

PR standards/11223

Add -G to take a single group argument heading towards POSIX.2 compliance.
Patch from jperkin and reviewed by simonb.

branches: 1.93.6;
ps(1): Guard freeing the memory of pinfo with __NO_LEAKS.

No more leaks are detected by LSan/NetBSD as of the LLVM snapshot
(clang10svn) from 2019-09-15.

Plug memory leak in ps(1)

pinfo is allocated in setpinfo() with calloc(3).

Free it when no longer used, just before the program termination.

Detected with LSan.

branches: 1.91.2;
kvm_geterr() already contains errno, use errx.

use EXIT_FAILURE instead of 1

branches: 1.89.2;
Fix an unitialized memory read bug in ps(1)

rawcpu of type int, is declared inside main(){} and it can be passed as
uninitialized to setpinfo().
The setpinfo() function has a switch checking the value of rawcpu:

if (!rawcpu)
pi[i].pcpu /= 1.0 - exp(ki[i].p_swtime * log_ccpu);

rawcpu is set to 1 with the command line argument "-C".

-C Change the way the CPU percentage is calculated by using a
"raw" CPU calculation that ignores "resident" time (this
normally has no effect).

Bug reproducible with an invocation: "ps u". It hides with "ps uC".

Initialize rawcpu by default to 0, before the getopt(3) machinery.

Detected with MSan running on NetBSD/amd64.

Sponsored by <The NetBSD Foundation>

branches: 1.88.6;
simplify logic; there must be no CPU usage when p_swtime is zero

provide a tree like display with -d, from FreeBSD

Calculate CPU usage (pcpu) once per process if it is required. This change
significantly improves performance for slow machines when output is sorted
by pcpu.

ok martin

KNF
ok martin

Drop main() prototype.

branches: 1.83.2;
remove \n from error messages

Don't print junk errnos. Avoids:
% ps -tfoo
ps: /dev/ttyfoo: Undefined error: 0
Exit 1

Don't bother using variables whose value is never changed from the
initialization value.

branches: 1.80.2;
Remove some pointless inclusions os sys/user.h

Make ps -s use LTIME instead of TIME in the default output format.

branches: 1.78.2;
Push logic to convert a ttyname to a device number into its own
function. Improve dealing with ptyfs by explicitly handling missing
pts/%d entries, if the kernel supports the pts device (PR 40813).

Choose better argument name for `-U' (sync with man page).
From Bug Hunting in email.

branches: 1.76.2; 1.76.4;
Use __dead

Fix ps(1) and top(1) to show reasonable CPU numbers i.e. cpu_index() provided
by the kernel, instead of CPU order number, which is generally random.

- add new RLIMIT_AS (aka RLIMIT_VMEM) resource that limits the total
address space available to processes. this limit exists in most other
modern unix variants, and like most of them, our defaults are unlimited.
remove the old mmap / rlimit.datasize hack.

- adds the VMCMD_STACK flag to all the stack-creation vmcmd callers.
it is currently unused, but was added a few years ago.

- add a pair of new process size values to kinfo_proc2{}. one is the
total size of the process memory map, and the other is the total size
adjusted for unused stack space (since most processes have a lot of
this...)

- patch sh, and csh to notice RLIMIT_AS. (in some cases, the alias
RLIMIT_VMEM was already present and used if availble.)

- patch ps, top and systat to notice the new k_vm_vsize member of
kinfo_proc2{}.

- update irix, svr4, svr4_32, linux and osf1 emulations to support
this information. (freebsd could be done, but that it's best left
as part of the full-update of compat/freebsd.)


this addresses PR 7897. it also gives correct memory usage values,
which have never been entirely correct (since mmap), and have been
very incorrect since jemalloc() was enabled.

tested on i386 and sparc64, build tested on several other platforms.

thanks to many folks for feedback and testing but most espcially
chuq and yamt for critical suggestions that lead to this patch not
having a special ugliness i wasn't happy with anyway :-)

fix -Wsign-compare issues

branches: 1.72.2;
fix dev_t 64 bit issues.

branches: 1.71.4;
Sync usage with man page.

Add "ps -A" option. This displays information about all processes,
like BSD "ps -ax" or SYSV "ps -ef". It's specified in SUSv3.

Also make some minor clarifications to the man page.

Remove the \n and tabs from the __COPYRIGHT() strings.

PR/38755: murray armfield: /bin posix programs missing setprogname and/or
setlocale

branches: 1.67.2;
Remove clause 3 and 4 from TNF licenses

branches: 1.66.2;
ps -t? needs to imply -x. Otherwise the non-x logic filters all the results
out. Thanks to der Mouse for prompting me to fix it...

branches: 1.65.2;
Add a CPUID keyword; idea and some of the code from Anon Ymous

make -s alter the default format rather than doing parsefmt.
it allows users to use -o to replace it later.

branches: 1.63.4;
Revert the changes done after newlock2 merge, except for the removal of
the now unused LSDEAD state. The current kernel headers are now
source-compatible with pre-newlock2 kernel headers again.

Approved by Andrew Doran.

Sync with kernel changes introduced by merging the newlock2 branch.

* When all columns are given null customised headers, the blank header
line is not printed at all. This is specified in P1003.1-2004
(SUSv3), and is useful.

* Customised headers may contain embedded space, commas and equals
signs. To specify multiple customised headers, use multiple -o or -O
options. This is specified (for "-o", not for "-O") in P1003.1-2004
(SUSv3), and is useful.

* When a column is given a null (blank) customised header, it keeps its
default minimum width. This is specified in P1003.1-2004 (SUSv3), and
seems harmless.

* Fix a bug that made it impossible to print the same keyword multiple
times, with different customised headers each time. (Previously, the
last customised header was used for all instances of the keyword.)

* Make the behaviour of "-O" more useful. The first -O option adds
the default keywords only if there have not yet been any formatting
options, and multiple -O options now insert their keywords in adjacent
positions. Now {ps -j -O %cpu} is like {ps -j} with one extra column;
Previously, it would have had all the columns implied by "-j", plus
all the default columns, plus the extra column specified by "-O".

* Convert from home-grown linked lists to SIMPLEQ lists.

Discussed in tech-userlevel.

factor out the code that does numeric parsing into a function and check for
errors properly.

sprinkle a little const, and now everything compiles with WARNS=3

appease gcc -Wuninitialized

allow -t ttyp0 -t p0 -t /dev/ttyp0 -t /dev/pts/0

Make ps understand "-tpts/4" by assuming that ptyfs is mounted on
/dev/pts and that the '/' immediately preceding the digits does not
need a 'p' in front of it (ick, kludge_oldps_options()).

Add (unsigned char) cast to ctype function

#define<TAB>

Wrap some long lines.
Consisently use (void) in front of printf(...);
A few more white-space nits.

KNF, ANSFify, de-__P, unC++ify, white space nits.

g/c SDEAD

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22249, verified by myself.

- add keywords for utime, stime and ctime (output sss.ssssss if small)
- fiddle with keyword table - mainly to keep width under 80 cols)
- add -k keylist so sort output
- don't use logs to work out field width, get -MAXINT right
- user defined headers can be 1 char (actually they can be zero length)
so ?? in tty column isn't guaranteed to be wider than header
- use PRId64 to remove some long long casts
- use kvm_openfiles(NULL, NULL, NULL, KVN_NO_FILES, ..) unless core file
specified
(approved by christos)

Merge the nathanw_sa branch.

g/c procfs fallback code, it's redundant now that we use sysctl to get
process info

branches: 1.46.2;
Sync getopt() / man page with actual getopt options.

fix minor WARNS=2

Remove duplicate `if'. Noted by Geoff C. Wing in bin/13551.

Add a small optimizetion in savevars(). break out of the loop once needcomm
is set since setting it multiple times only wastes cpu time.

If there were no processes to list, it meant that our criteria failed
and didnt match anything. Dont make our caller think we succeeded.

Don't check namelist when choosing to use the sysctl interface.

Fix handling of "ps U<user>" where <user> ended it 't' - now if the
first argument doesn't start with a '-' and ends in 't', the 't' only
gets converted to a 'T' (for listing processes on the current tty)
if it doesn't contain any other command line option characters that
take an argument. Problem noticed by ITOH Yasufumi on tech-userlevel.

If the 'start' format was the last column, the width wouldn't be set
and started() output a zero length string instead of a string the right
length - fix header width calculations so the the last column is handled
the same all others. This only showed up for 'start' because started()
knew that the column header is the same width and the field itself...

Update copyrights.

Calculate field widths on the fly so that all columns line up nicely
and make more effective use of screen real estate when some columns
(eg USER and VSZ) didn't need the full default width.

branches: 1.38.2;
Use new sysctl/kvm interfaces. This will stop the "proc size mismatch"
errors when internal kernel structures change size. Also remove
the sgid bit - all live kernel data is accessed through the sysctl
interface.

If ps cannot get the terminal size from STDOUT, it tried STDERR and STDIN.
What this does to the POLS when a script that is parsing ps output just happens
to run in a narrow terminal cannot be described in polite company.
If STDOUT is not a tty, leave width at 80.

Use tty "?" to select processes with no controlling tty and tty "-" to
select processes with a revoked controlling tty.

Idea from SunOS/Solaris, suggested by Alexis Rosen.

make -M imply -K. the current system is not likely to be
a reasonable approximation of a crash dump.

Document -U switch (displays processes belonging to given user/uid).
Reported in PR 8943 by WL <wliao@dura.spc.uchicago.edu>

"ps -e" only shows the environment for the processes owned by the
current user id or all process if run by root.

Fixes PR security/5967 from Todd Vierling.

Since our gcc doesn't warn about NULL format strings anymore, we can
fix the incorrect err(1, "%s", "") et al.
Closes PR bin/7592 by cgd.

command(): don't start using procfs_getargv() instead of kvm_getargv()
just because the latter returned a NULL - that may normally happen,
when the process is, for example, a kernel thread

procfs_getargv(): if the name is brackened in '(' and ')', just
return NULL immediately - the code in command() will DTRT
and the output will be same for procfs- and kvm- based lookup
in all cases

getkinfo_kvm(): fix bug in previous - handled size_t i as if it would be signed
saveuser(): don't check kd - if we get here, kd is never NULL

When using the procfs for extracting process information, extract
also process start time, process arguments and session leadership status.

The procfs fallback is also used when kvm_openfiles() completely fails
(e.g. when /dev/mem is not readable).

Solves PR 7772, though the final implementation is different.

branches: 1.28.2; 1.28.4;
Minor fixes:
1. Use statfs() to verify that /proc is a procfs.
2. Add -K option to disable /proc-based method.
3. Make warnings less verbose, but still accurate.

Added experimental fallback /proc-based lookup. Warnings are printed when the
fallback method is used, as the results could be untrustworthy if an
intruder is present. It is highly likely that NetBSD-1.5 will have
an improved kvm interface for reading process information, at which
point this code can be garbage-collected. Also added a word to the
man page -x option description while I was here.

Fixed segfaulting for "ps -T". There is no optarg with -T, so instructions
under label "tty:" need to use ttname instead.

Add a -U option to select by user/uid.

Be more retentive about use of NOTREACHED and noreturn.

Delint.

__AUDIT__ cleanup.

- change setgid kmem programs (that lend themselves to this) so setegid(getgid())
and the top, and then set the effective gid back to kmem around the call to
kvm_openfiles(). this reduces the time group kmem is available.
- for those above that also allow this, setgid(getgid()) after the call to
kvm_openfiles() to fully revoke priviledges.
- some KNF
- use err(3) over fprintf(3) in some places

getopt returns -1 not EOF
cleanup manpage

Fix compiler warnings
Add WARNS=1

note that strcpy is safe.

Message buffer passed to kvm_openfiles() should _POSIX2_LINE_MAX long; PR#3266

Fix SEGV when kludging old-style options.
From Greg Stark <gsstark@mit.edu>, PR #2550.

Create and strvis(3) the argument and environment lists only as needed.
Do not keep extra copies around.

Gather the argument and environment lists when we actually want to print them.

convert to new RCS id conventions.

Integrate `-c' option, from Arne Juul.

clean up import

update for changed flags

fix from Thomas Eberhardt <thomas@mathematik.uni-Bremen.de>:
don't try to display the resident text size; we can't do that anymore

Add RCS identifiers.

branches: 1.7.2;
from Brian Somers <brian@awfulhak.demon.co.uk>:
The kvm entry isn't updated for command unless 'time' is given,
therefore
ps -o command
outputs blank lines [or, in NetBSD, seg faults]
ps -o time,command
works.

fix alien 'w' handling. you really don't want to know...

add support and documentation for the 'e' option, to print out
the process environment.

changed "Id" to "Header" for rcsids

added rcs ids to all files

after 0.2.2 "stable" patches applied

branches: 1.1.1;
Initial revision

use emalloc and friends, add ktrace flag printing.

Provide symbolic printing of some keywords by capitalizing them.

bin: remove unnecessary lint comment CONSTCOND

Since 2021-01-31, lint no longer warns about 'do ... while (0)'.

No functional change.

branches: 1.29.14;
provide a tree like display with -d, from FreeBSD

Calculate CPU usage (pcpu) once per process if it is required. This change
significantly improves performance for slow machines when output is sorted
by pcpu.

ok martin

branches: 1.27.6;
Use an enum type for PRINTMODE vs. WIDTHMODE. Compiler output diffs have
been checked.

branches: 1.26.40; 1.26.46; 1.26.60;
* When all columns are given null customised headers, the blank header
line is not printed at all. This is specified in P1003.1-2004
(SUSv3), and is useful.

* Customised headers may contain embedded space, commas and equals
signs. To specify multiple customised headers, use multiple -o or -O
options. This is specified (for "-o", not for "-O") in P1003.1-2004
(SUSv3), and is useful.

* When a column is given a null (blank) customised header, it keeps its
default minimum width. This is specified in P1003.1-2004 (SUSv3), and
seems harmless.

* Fix a bug that made it impossible to print the same keyword multiple
times, with different customised headers each time. (Previously, the
last customised header was used for all instances of the keyword.)

* Make the behaviour of "-O" more useful. The first -O option adds
the default keywords only if there have not yet been any formatting
options, and multiple -O options now insert their keywords in adjacent
positions. Now {ps -j -O %cpu} is like {ps -j} with one extra column;
Previously, it would have had all the columns implied by "-j", plus
all the default columns, plus the extra column specified by "-O".

* Convert from home-grown linked lists to SIMPLEQ lists.

Discussed in tech-userlevel.

sprinkle a little const, and now everything compiles with WARNS=3

Add support for the POSIX.2 "args" and "comm" keywords. "args" is just
an alias for "command", and "comm" shows just argv[0] and not the whole
argument list.

Fix for part of PR standards/11224.

#define<TAB>

KNF, ANSFify, de-__P, unC++ify, white space nits.

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22249, verified by myself.

- add keywords for utime, stime and ctime (output sss.ssssss if small)
- fiddle with keyword table - mainly to keep width under 80 cols
- add -k keylist so sort output
- don't use logs to work out field width, get -MAXINT right
- user defined headers can be 1 char (actually they can be zero length)
so ?? in tty column isn't guaranteed to be wider than header
- remove some long long casts (compiles on sparc64 still)
- use kvm_openfiles(NULL, NULL, NULL, KVN_NO_FILES, ..) unless core file
specified,
(approved by christos)

Merge the nathanw_sa branch.

Since kinfo_proc2 has many values that are 64-bit, ps needs to know
that they're 64-bit, and grab them out of memory appropriately. Otherwise,
big-endian systems get the wrong end of the 64-bit value and lose.

Keywords affected: inblk, majflt, minflt, msgrcv, msgsnd, nivcsw, nsigs,
nswap, nvcsw, and oublk.

branches: 1.17.4;
Calculate field widths on the fly so that all columns line up nicely
and make more effective use of screen real estate when some columns
(eg USER and VSZ) didn't need the full default width.

branches: 1.16.2;
Use new sysctl/kvm interfaces. This will stop the "proc size mismatch"
errors when internal kernel structures change size. Also remove
the sgid bit - all live kernel data is accessed through the sysctl
interface.

oops, fix merge botch in previous..

limit session to 6 chars. from Geoff Wing <gcw@pobox.com> and fixes his PR#7506

Add support for printing wide sigset_t properly.

Fix p_nice vs. NZERO code.
Remove some unnecessary casts.
Remove an unused time format field.

add INT32 and UINT32 variable type specifiers. some of the variables
that are printed are of fixed size.

Kenneth Stailey pointed out that his previous fix was incorrect. RE and SL
should be bounded at 127, not 99.

as suggested in PR 1347, by Kenneth Stailey, make "RE" and "SL" fields
show up as two digits, max, displaying "99" if > 99. allow any field with
an unsigned value to be specified as printing that way (though right now
it's only used for RE and SL, which are both unsigned longs). This fix
is substantially different than that suggested in the PR.

Create and strvis(3) the argument and environment lists only as needed.
Do not keep extra copies around.

add the ability to understand and print INT and UINT-sized variables.

convert to new RCS id conventions.

clean up import

Add RCS identifiers.

changed "Id" to "Header" for rcsids

added rcs ids to all files

branches: 1.1.1;
Initial revision