Replace use of the old BSD st_*timespec members in struct stat with
the POSIX-standard st_*tim members.

ok millert@

horrible whitespace, mostly on non-code lines. no object change

Delete obsolete /* ARGSUSED */ lint comments.

ok miod@ millert@

Delete obsolete /* ARGSUSED1 */ lint comments.

ok claudio@ cheloha@ krw@ deraadt@ miod@ millert@

spelling fixes from paul tagliamonte;
part of a larger diff: i'm committing this separately
as i changed statuses to status;

The -exec primary is terminated by either ';' or '+', but the latter
only if immediately following a '{}' placeholder. Slightly modify the
error message to indicate so.

Brought up by and discussed with Paul de Weerd, thanks!

ok millert@

Account for environment when computing maximum space used for args

Already done by xargs(1). Pointed out by espie@, ok millert@

find -exec +: use sysconf to find the kernel's idea of ARG_MAX

Using ARG_MAX directly doesn't fly when ARG_MAX gets bumped and the
kernel and userland are not in sync, effectively breaking find -exec +.
Use sysconf(3) as already done in xargs(1).

Spotted by sthen@, ok deraadt@ millert@ sthen@

When system calls indicate an error they return -1, not some arbitrary
value < 0. errno is only updated in this case. Change all (most?)
callers of syscalls to follow this better, and let's see if this strictness
helps us in the future.

Use uid_from_user(3) and gid_from_group(3) in utilities that
do repeated lookups. OK tb@

add -delete option which can simplify the common case of wanting to delete
lots of files without the arcane -exec or error prone xargs.
code from freebsd.
ok millert

Convert many atoi() calls to strtonum(), adding range checks and failure
handling along the way.
Reviews by Brendan MacDonell, Jeremy Devenport, florian, doug, millert

tzfile.h is an internal header that should never have been installed.
What's worse, the tzfile.h that gets installed is over 20 years old
and doesn't match the real tzfile.h in libc/time. This makes the
tree safe for /usr/include/tzfile.h removal. The TM_YEAR_BASE
define has been moved to time.h temporarily until its usage is
replaced by 1900 in the tree. Actual removal of tzfile.h is pending
a ports build. Based on a diff from deraadt@

Replace <sys/param.h> with <limits.h> and other less dirty headers where
possible. Annotate <sys/param.h> lines with their current reasons. Switch
to PATH_MAX, NGROUPS_MAX, HOST_NAME_MAX+1, LOGIN_NAME_MAX, etc. Change
MIN() and MAX() to local definitions of MINIMUM() and MAXIMUM() where
sensible to avoid pulling in the pollution. These are the files confirmed
through binary verification.
ok guenther, millert, doug (helped with the verification protocol)

use reallocarray
okay chl@, tedu@

adjust comment

handle large numbers. this fixes time_t beyond 2038, constrains the
range of i_num correctly, and now handles files > 4GB in size on 32-bit
machines.
ok otto millert

Fix typos in comment (s,exacty,exactly,)

Add support for "find ... -exec foo {} \+".

From the manpage:

If terminated by a plus sign (``+''), the pathnames for which the
primary is evaluated are aggregated into sets, and utility will
be invoked once per set, similar to xargs(1). If any invocation
exits with non-zero exit status, then find will eventually do so
as well, but this does not cause find to exit early. The string
``{}'' must appear, and must appear last. Each set is limited to
no more than 5,000 pathnames, and is also limited such that the
invocation of utility does not exceed ARG_MAX.

Code comes from NetBSD, written by John Hawkinson. Tested by eivinde at
terraplane dot org and myself for quite some time.

ok miod@

Don't error out if we can't opendir() when -empty is used.
Fix by Bakul Shah. OK deraadt@ halex@

We only need to fchdir(dotfd) for the -exec and -ok primaries so
defer the check for dotfd != -1 until we know we will need it.
Based on a diff from schwarze@

rcsid[] and sccsid[] and copyright[] are essentially unmaintained (and
unmaintainable). these days, people use source. these id's do not provide
any benefit, and do hurt the small install media
(the 33,000 line diff is essentially mechanical)
ok with the idea millert, ok dms

Allow -maxdepth 0; noticed by bluhm@

Don't stop traversing a directory hierarchy if we reach SHRT_MAX,
just stop updating fts_level so we don't overflow it. This allows
rm, find, etc to operate on very deep hierarchies. Consumers of
fts(3) do need to be aware that the actual level may be larger
than fts_level. During the next libc major bump we will make
fts_level an int instead of a short. OK deraadt@

no more whiteouts

ansi; khalek@linuxgamers.net

setmode(3) returns void *, not mode_t *. ok millert@

bring protypes into scope. this requires some quirky handling, but in
the end everything is much clearer; ok tedu (itojun might like to see
how ifconfig looks after this)

mostly ansi cleanup; pval ok

Remove the advertising clause in the UCB license which Berkeley
rescinded 22 July 1999. Proofed by myself and Theo.

remove local declaration of user_from_uid()/group_from_uid()
ok millert@

Fix pasto in -anewer and -cnewer options that caused them to behave
as -newer; Dave Horsfall

Manual cleanup of remaining userland __P use (excluding packages maintained outside the tree)

Part one of userland __P removal. Done with a simple regexp with some minor hand editing to make comments line up correctly. Another pass is forthcoming that handles the cases that could not be done automatically.

kill more registers

millert@ ok

new strtofflags/fflagstostr

Add a -flags option similar to -perm but for filke flags (ala
chflags). Adapted from FreeBSD.

-iname support; karls@inet.no

o use nanosecond precision with the -newer option
o add -anewer and -cnewer options like GNU find.

If we can't chdir to the root dir in -execdir, just return instead
of generating an error. Most times this is just because the target
is not a directory. Also close an fd leak introduced in the last
commit.

Move the chdir/fchdir hack so that it is local to the -execdir
function, since that's the only place it needs to be. Doing it for
each function caused problems with multiple directories.

Do not use FTS_CHDIRROOT flag as it is fatally flawed. Instead,
do the chdir ourselves. This fixes a problem with the -execdir option
with multiple relative directories.

Add -W option so -type W gets a chance to work.

more setmode() leaks -- kill 'em all

Add the primaries -mmin, -amin, -cmin to find, similar to the GNU find; wosch

Fix problem with ``find -execdir'' not having the correct initial cwd.
Adds a new flag to fts(3).

Some -Wall cleanliness.

Add -execdir support.

Add support for -empty

Add support for -mindepth while at it

Add support for -maxdepth option

rcsid

indent nice

support whiteouts; netbsd pr#2394; mike.long@analog.com

branches: 1.1.1;
Initial revision

horrible whitespace, mostly on non-code lines. no object change

Delete obsolete /* ARGSUSED */ lint comments.

ok miod@ millert@

Delete obsolete /* ARGSUSED1 */ lint comments.

ok claudio@ cheloha@ krw@ deraadt@ miod@ millert@

spelling fixes from paul tagliamonte;
part of a larger diff: i'm committing this separately
as i changed statuses to status;

The -exec primary is terminated by either ';' or '+', but the latter
only if immediately following a '{}' placeholder. Slightly modify the
error message to indicate so.

Brought up by and discussed with Paul de Weerd, thanks!

ok millert@

Account for environment when computing maximum space used for args

Already done by xargs(1). Pointed out by espie@, ok millert@

find -exec +: use sysconf to find the kernel's idea of ARG_MAX

Using ARG_MAX directly doesn't fly when ARG_MAX gets bumped and the
kernel and userland are not in sync, effectively breaking find -exec +.
Use sysconf(3) as already done in xargs(1).

Spotted by sthen@, ok deraadt@ millert@ sthen@

When system calls indicate an error they return -1, not some arbitrary
value < 0. errno is only updated in this case. Change all (most?)
callers of syscalls to follow this better, and let's see if this strictness
helps us in the future.

Use uid_from_user(3) and gid_from_group(3) in utilities that
do repeated lookups. OK tb@

add -delete option which can simplify the common case of wanting to delete
lots of files without the arcane -exec or error prone xargs.
code from freebsd.
ok millert

Convert many atoi() calls to strtonum(), adding range checks and failure
handling along the way.
Reviews by Brendan MacDonell, Jeremy Devenport, florian, doug, millert

tzfile.h is an internal header that should never have been installed.
What's worse, the tzfile.h that gets installed is over 20 years old
and doesn't match the real tzfile.h in libc/time. This makes the
tree safe for /usr/include/tzfile.h removal. The TM_YEAR_BASE
define has been moved to time.h temporarily until its usage is
replaced by 1900 in the tree. Actual removal of tzfile.h is pending
a ports build. Based on a diff from deraadt@

Replace <sys/param.h> with <limits.h> and other less dirty headers where
possible. Annotate <sys/param.h> lines with their current reasons. Switch
to PATH_MAX, NGROUPS_MAX, HOST_NAME_MAX+1, LOGIN_NAME_MAX, etc. Change
MIN() and MAX() to local definitions of MINIMUM() and MAXIMUM() where
sensible to avoid pulling in the pollution. These are the files confirmed
through binary verification.
ok guenther, millert, doug (helped with the verification protocol)

use reallocarray
okay chl@, tedu@

adjust comment

handle large numbers. this fixes time_t beyond 2038, constrains the
range of i_num correctly, and now handles files > 4GB in size on 32-bit
machines.
ok otto millert

Fix typos in comment (s,exacty,exactly,)

Add support for "find ... -exec foo {} \+".

From the manpage:

If terminated by a plus sign (``+''), the pathnames for which the
primary is evaluated are aggregated into sets, and utility will
be invoked once per set, similar to xargs(1). If any invocation
exits with non-zero exit status, then find will eventually do so
as well, but this does not cause find to exit early. The string
``{}'' must appear, and must appear last. Each set is limited to
no more than 5,000 pathnames, and is also limited such that the
invocation of utility does not exceed ARG_MAX.

Code comes from NetBSD, written by John Hawkinson. Tested by eivinde at
terraplane dot org and myself for quite some time.

ok miod@

Don't error out if we can't opendir() when -empty is used.
Fix by Bakul Shah. OK deraadt@ halex@

We only need to fchdir(dotfd) for the -exec and -ok primaries so
defer the check for dotfd != -1 until we know we will need it.
Based on a diff from schwarze@

rcsid[] and sccsid[] and copyright[] are essentially unmaintained (and
unmaintainable). these days, people use source. these id's do not provide
any benefit, and do hurt the small install media
(the 33,000 line diff is essentially mechanical)
ok with the idea millert, ok dms

Allow -maxdepth 0; noticed by bluhm@

Don't stop traversing a directory hierarchy if we reach SHRT_MAX,
just stop updating fts_level so we don't overflow it. This allows
rm, find, etc to operate on very deep hierarchies. Consumers of
fts(3) do need to be aware that the actual level may be larger
than fts_level. During the next libc major bump we will make
fts_level an int instead of a short. OK deraadt@

no more whiteouts

ansi; khalek@linuxgamers.net

setmode(3) returns void *, not mode_t *. ok millert@

bring protypes into scope. this requires some quirky handling, but in
the end everything is much clearer; ok tedu (itojun might like to see
how ifconfig looks after this)

mostly ansi cleanup; pval ok

Remove the advertising clause in the UCB license which Berkeley
rescinded 22 July 1999. Proofed by myself and Theo.

remove local declaration of user_from_uid()/group_from_uid()
ok millert@

Fix pasto in -anewer and -cnewer options that caused them to behave
as -newer; Dave Horsfall

Manual cleanup of remaining userland __P use (excluding packages maintained outside the tree)

Part one of userland __P removal. Done with a simple regexp with some minor hand editing to make comments line up correctly. Another pass is forthcoming that handles the cases that could not be done automatically.

kill more registers

millert@ ok

new strtofflags/fflagstostr

Add a -flags option similar to -perm but for filke flags (ala
chflags). Adapted from FreeBSD.

-iname support; karls@inet.no

o use nanosecond precision with the -newer option
o add -anewer and -cnewer options like GNU find.

If we can't chdir to the root dir in -execdir, just return instead
of generating an error. Most times this is just because the target
is not a directory. Also close an fd leak introduced in the last
commit.

Move the chdir/fchdir hack so that it is local to the -execdir
function, since that's the only place it needs to be. Doing it for
each function caused problems with multiple directories.

Do not use FTS_CHDIRROOT flag as it is fatally flawed. Instead,
do the chdir ourselves. This fixes a problem with the -execdir option
with multiple relative directories.

Add -W option so -type W gets a chance to work.

more setmode() leaks -- kill 'em all

Add the primaries -mmin, -amin, -cmin to find, similar to the GNU find; wosch

Fix problem with ``find -execdir'' not having the correct initial cwd.
Adds a new flag to fts(3).

Some -Wall cleanliness.

Add -execdir support.

Add support for -empty

Add support for -mindepth while at it

Add support for -maxdepth option

rcsid

indent nice

support whiteouts; netbsd pr#2394; mike.long@analog.com

branches: 1.1.1;
Initial revision

Delete obsolete /* ARGSUSED */ lint comments.

ok miod@ millert@

Delete obsolete /* ARGSUSED1 */ lint comments.

ok claudio@ cheloha@ krw@ deraadt@ miod@ millert@

spelling fixes from paul tagliamonte;
part of a larger diff: i'm committing this separately
as i changed statuses to status;

The -exec primary is terminated by either ';' or '+', but the latter
only if immediately following a '{}' placeholder. Slightly modify the
error message to indicate so.

Brought up by and discussed with Paul de Weerd, thanks!

ok millert@

Account for environment when computing maximum space used for args

Already done by xargs(1). Pointed out by espie@, ok millert@

find -exec +: use sysconf to find the kernel's idea of ARG_MAX

Using ARG_MAX directly doesn't fly when ARG_MAX gets bumped and the
kernel and userland are not in sync, effectively breaking find -exec +.
Use sysconf(3) as already done in xargs(1).

Spotted by sthen@, ok deraadt@ millert@ sthen@

When system calls indicate an error they return -1, not some arbitrary
value < 0. errno is only updated in this case. Change all (most?)
callers of syscalls to follow this better, and let's see if this strictness
helps us in the future.

Use uid_from_user(3) and gid_from_group(3) in utilities that
do repeated lookups. OK tb@

add -delete option which can simplify the common case of wanting to delete
lots of files without the arcane -exec or error prone xargs.
code from freebsd.
ok millert

Convert many atoi() calls to strtonum(), adding range checks and failure
handling along the way.
Reviews by Brendan MacDonell, Jeremy Devenport, florian, doug, millert

tzfile.h is an internal header that should never have been installed.
What's worse, the tzfile.h that gets installed is over 20 years old
and doesn't match the real tzfile.h in libc/time. This makes the
tree safe for /usr/include/tzfile.h removal. The TM_YEAR_BASE
define has been moved to time.h temporarily until its usage is
replaced by 1900 in the tree. Actual removal of tzfile.h is pending
a ports build. Based on a diff from deraadt@

Replace <sys/param.h> with <limits.h> and other less dirty headers where
possible. Annotate <sys/param.h> lines with their current reasons. Switch
to PATH_MAX, NGROUPS_MAX, HOST_NAME_MAX+1, LOGIN_NAME_MAX, etc. Change
MIN() and MAX() to local definitions of MINIMUM() and MAXIMUM() where
sensible to avoid pulling in the pollution. These are the files confirmed
through binary verification.
ok guenther, millert, doug (helped with the verification protocol)

use reallocarray
okay chl@, tedu@

adjust comment

handle large numbers. this fixes time_t beyond 2038, constrains the
range of i_num correctly, and now handles files > 4GB in size on 32-bit
machines.
ok otto millert

Fix typos in comment (s,exacty,exactly,)

Add support for "find ... -exec foo {} \+".

From the manpage:

If terminated by a plus sign (``+''), the pathnames for which the
primary is evaluated are aggregated into sets, and utility will
be invoked once per set, similar to xargs(1). If any invocation
exits with non-zero exit status, then find will eventually do so
as well, but this does not cause find to exit early. The string
``{}'' must appear, and must appear last. Each set is limited to
no more than 5,000 pathnames, and is also limited such that the
invocation of utility does not exceed ARG_MAX.

Code comes from NetBSD, written by John Hawkinson. Tested by eivinde at
terraplane dot org and myself for quite some time.

ok miod@

Don't error out if we can't opendir() when -empty is used.
Fix by Bakul Shah. OK deraadt@ halex@

We only need to fchdir(dotfd) for the -exec and -ok primaries so
defer the check for dotfd != -1 until we know we will need it.
Based on a diff from schwarze@

rcsid[] and sccsid[] and copyright[] are essentially unmaintained (and
unmaintainable). these days, people use source. these id's do not provide
any benefit, and do hurt the small install media
(the 33,000 line diff is essentially mechanical)
ok with the idea millert, ok dms

Allow -maxdepth 0; noticed by bluhm@

Don't stop traversing a directory hierarchy if we reach SHRT_MAX,
just stop updating fts_level so we don't overflow it. This allows
rm, find, etc to operate on very deep hierarchies. Consumers of
fts(3) do need to be aware that the actual level may be larger
than fts_level. During the next libc major bump we will make
fts_level an int instead of a short. OK deraadt@

no more whiteouts

ansi; khalek@linuxgamers.net

setmode(3) returns void *, not mode_t *. ok millert@

bring protypes into scope. this requires some quirky handling, but in
the end everything is much clearer; ok tedu (itojun might like to see
how ifconfig looks after this)

mostly ansi cleanup; pval ok

Remove the advertising clause in the UCB license which Berkeley
rescinded 22 July 1999. Proofed by myself and Theo.

remove local declaration of user_from_uid()/group_from_uid()
ok millert@

Fix pasto in -anewer and -cnewer options that caused them to behave
as -newer; Dave Horsfall

Manual cleanup of remaining userland __P use (excluding packages maintained outside the tree)

Part one of userland __P removal. Done with a simple regexp with some minor hand editing to make comments line up correctly. Another pass is forthcoming that handles the cases that could not be done automatically.

kill more registers

millert@ ok

new strtofflags/fflagstostr

Add a -flags option similar to -perm but for filke flags (ala
chflags). Adapted from FreeBSD.

-iname support; karls@inet.no

o use nanosecond precision with the -newer option
o add -anewer and -cnewer options like GNU find.

If we can't chdir to the root dir in -execdir, just return instead
of generating an error. Most times this is just because the target
is not a directory. Also close an fd leak introduced in the last
commit.

Move the chdir/fchdir hack so that it is local to the -execdir
function, since that's the only place it needs to be. Doing it for
each function caused problems with multiple directories.

Do not use FTS_CHDIRROOT flag as it is fatally flawed. Instead,
do the chdir ourselves. This fixes a problem with the -execdir option
with multiple relative directories.

Add -W option so -type W gets a chance to work.

more setmode() leaks -- kill 'em all

Add the primaries -mmin, -amin, -cmin to find, similar to the GNU find; wosch

Fix problem with ``find -execdir'' not having the correct initial cwd.
Adds a new flag to fts(3).

Some -Wall cleanliness.

Add -execdir support.

Add support for -empty

Add support for -mindepth while at it

Add support for -maxdepth option

rcsid

indent nice

support whiteouts; netbsd pr#2394; mike.long@analog.com

branches: 1.1.1;
Initial revision

spelling fixes from paul tagliamonte;
part of a larger diff: i'm committing this separately
as i changed statuses to status;

The -exec primary is terminated by either ';' or '+', but the latter
only if immediately following a '{}' placeholder. Slightly modify the
error message to indicate so.

Brought up by and discussed with Paul de Weerd, thanks!

ok millert@

Account for environment when computing maximum space used for args

Already done by xargs(1). Pointed out by espie@, ok millert@

find -exec +: use sysconf to find the kernel's idea of ARG_MAX

Using ARG_MAX directly doesn't fly when ARG_MAX gets bumped and the
kernel and userland are not in sync, effectively breaking find -exec +.
Use sysconf(3) as already done in xargs(1).

Spotted by sthen@, ok deraadt@ millert@ sthen@

When system calls indicate an error they return -1, not some arbitrary
value < 0. errno is only updated in this case. Change all (most?)
callers of syscalls to follow this better, and let's see if this strictness
helps us in the future.

Use uid_from_user(3) and gid_from_group(3) in utilities that
do repeated lookups. OK tb@

add -delete option which can simplify the common case of wanting to delete
lots of files without the arcane -exec or error prone xargs.
code from freebsd.
ok millert

Convert many atoi() calls to strtonum(), adding range checks and failure
handling along the way.
Reviews by Brendan MacDonell, Jeremy Devenport, florian, doug, millert

tzfile.h is an internal header that should never have been installed.
What's worse, the tzfile.h that gets installed is over 20 years old
and doesn't match the real tzfile.h in libc/time. This makes the
tree safe for /usr/include/tzfile.h removal. The TM_YEAR_BASE
define has been moved to time.h temporarily until its usage is
replaced by 1900 in the tree. Actual removal of tzfile.h is pending
a ports build. Based on a diff from deraadt@

Replace <sys/param.h> with <limits.h> and other less dirty headers where
possible. Annotate <sys/param.h> lines with their current reasons. Switch
to PATH_MAX, NGROUPS_MAX, HOST_NAME_MAX+1, LOGIN_NAME_MAX, etc. Change
MIN() and MAX() to local definitions of MINIMUM() and MAXIMUM() where
sensible to avoid pulling in the pollution. These are the files confirmed
through binary verification.
ok guenther, millert, doug (helped with the verification protocol)

use reallocarray
okay chl@, tedu@

adjust comment

handle large numbers. this fixes time_t beyond 2038, constrains the
range of i_num correctly, and now handles files > 4GB in size on 32-bit
machines.
ok otto millert

Fix typos in comment (s,exacty,exactly,)

Add support for "find ... -exec foo {} \+".

From the manpage:

If terminated by a plus sign (``+''), the pathnames for which the
primary is evaluated are aggregated into sets, and utility will
be invoked once per set, similar to xargs(1). If any invocation
exits with non-zero exit status, then find will eventually do so
as well, but this does not cause find to exit early. The string
``{}'' must appear, and must appear last. Each set is limited to
no more than 5,000 pathnames, and is also limited such that the
invocation of utility does not exceed ARG_MAX.

Code comes from NetBSD, written by John Hawkinson. Tested by eivinde at
terraplane dot org and myself for quite some time.

ok miod@

Don't error out if we can't opendir() when -empty is used.
Fix by Bakul Shah. OK deraadt@ halex@

We only need to fchdir(dotfd) for the -exec and -ok primaries so
defer the check for dotfd != -1 until we know we will need it.
Based on a diff from schwarze@

rcsid[] and sccsid[] and copyright[] are essentially unmaintained (and
unmaintainable). these days, people use source. these id's do not provide
any benefit, and do hurt the small install media
(the 33,000 line diff is essentially mechanical)
ok with the idea millert, ok dms

Allow -maxdepth 0; noticed by bluhm@

Don't stop traversing a directory hierarchy if we reach SHRT_MAX,
just stop updating fts_level so we don't overflow it. This allows
rm, find, etc to operate on very deep hierarchies. Consumers of
fts(3) do need to be aware that the actual level may be larger
than fts_level. During the next libc major bump we will make
fts_level an int instead of a short. OK deraadt@

no more whiteouts

ansi; khalek@linuxgamers.net

setmode(3) returns void *, not mode_t *. ok millert@

bring protypes into scope. this requires some quirky handling, but in
the end everything is much clearer; ok tedu (itojun might like to see
how ifconfig looks after this)

mostly ansi cleanup; pval ok

Remove the advertising clause in the UCB license which Berkeley
rescinded 22 July 1999. Proofed by myself and Theo.

remove local declaration of user_from_uid()/group_from_uid()
ok millert@

Fix pasto in -anewer and -cnewer options that caused them to behave
as -newer; Dave Horsfall

Manual cleanup of remaining userland __P use (excluding packages maintained outside the tree)

Part one of userland __P removal. Done with a simple regexp with some minor hand editing to make comments line up correctly. Another pass is forthcoming that handles the cases that could not be done automatically.

kill more registers

millert@ ok

new strtofflags/fflagstostr

Add a -flags option similar to -perm but for filke flags (ala
chflags). Adapted from FreeBSD.

-iname support; karls@inet.no

o use nanosecond precision with the -newer option
o add -anewer and -cnewer options like GNU find.

If we can't chdir to the root dir in -execdir, just return instead
of generating an error. Most times this is just because the target
is not a directory. Also close an fd leak introduced in the last
commit.

Move the chdir/fchdir hack so that it is local to the -execdir
function, since that's the only place it needs to be. Doing it for
each function caused problems with multiple directories.

Do not use FTS_CHDIRROOT flag as it is fatally flawed. Instead,
do the chdir ourselves. This fixes a problem with the -execdir option
with multiple relative directories.

Add -W option so -type W gets a chance to work.

more setmode() leaks -- kill 'em all

Add the primaries -mmin, -amin, -cmin to find, similar to the GNU find; wosch

Fix problem with ``find -execdir'' not having the correct initial cwd.
Adds a new flag to fts(3).

Some -Wall cleanliness.

Add -execdir support.

Add support for -empty

Add support for -mindepth while at it

Add support for -maxdepth option

rcsid

indent nice

support whiteouts; netbsd pr#2394; mike.long@analog.com

branches: 1.1.1;
Initial revision

The -exec primary is terminated by either ';' or '+', but the latter
only if immediately following a '{}' placeholder. Slightly modify the
error message to indicate so.

Brought up by and discussed with Paul de Weerd, thanks!

ok millert@

Account for environment when computing maximum space used for args

Already done by xargs(1). Pointed out by espie@, ok millert@

find -exec +: use sysconf to find the kernel's idea of ARG_MAX

Using ARG_MAX directly doesn't fly when ARG_MAX gets bumped and the
kernel and userland are not in sync, effectively breaking find -exec +.
Use sysconf(3) as already done in xargs(1).

Spotted by sthen@, ok deraadt@ millert@ sthen@

When system calls indicate an error they return -1, not some arbitrary
value < 0. errno is only updated in this case. Change all (most?)
callers of syscalls to follow this better, and let's see if this strictness
helps us in the future.

Use uid_from_user(3) and gid_from_group(3) in utilities that
do repeated lookups. OK tb@

add -delete option which can simplify the common case of wanting to delete
lots of files without the arcane -exec or error prone xargs.
code from freebsd.
ok millert

Convert many atoi() calls to strtonum(), adding range checks and failure
handling along the way.
Reviews by Brendan MacDonell, Jeremy Devenport, florian, doug, millert

tzfile.h is an internal header that should never have been installed.
What's worse, the tzfile.h that gets installed is over 20 years old
and doesn't match the real tzfile.h in libc/time. This makes the
tree safe for /usr/include/tzfile.h removal. The TM_YEAR_BASE
define has been moved to time.h temporarily until its usage is
replaced by 1900 in the tree. Actual removal of tzfile.h is pending
a ports build. Based on a diff from deraadt@

Replace <sys/param.h> with <limits.h> and other less dirty headers where
possible. Annotate <sys/param.h> lines with their current reasons. Switch
to PATH_MAX, NGROUPS_MAX, HOST_NAME_MAX+1, LOGIN_NAME_MAX, etc. Change
MIN() and MAX() to local definitions of MINIMUM() and MAXIMUM() where
sensible to avoid pulling in the pollution. These are the files confirmed
through binary verification.
ok guenther, millert, doug (helped with the verification protocol)

use reallocarray
okay chl@, tedu@

adjust comment

handle large numbers. this fixes time_t beyond 2038, constrains the
range of i_num correctly, and now handles files > 4GB in size on 32-bit
machines.
ok otto millert

Fix typos in comment (s,exacty,exactly,)

Add support for "find ... -exec foo {} \+".

From the manpage:

If terminated by a plus sign (``+''), the pathnames for which the
primary is evaluated are aggregated into sets, and utility will
be invoked once per set, similar to xargs(1). If any invocation
exits with non-zero exit status, then find will eventually do so
as well, but this does not cause find to exit early. The string
``{}'' must appear, and must appear last. Each set is limited to
no more than 5,000 pathnames, and is also limited such that the
invocation of utility does not exceed ARG_MAX.

Code comes from NetBSD, written by John Hawkinson. Tested by eivinde at
terraplane dot org and myself for quite some time.

ok miod@

Don't error out if we can't opendir() when -empty is used.
Fix by Bakul Shah. OK deraadt@ halex@

We only need to fchdir(dotfd) for the -exec and -ok primaries so
defer the check for dotfd != -1 until we know we will need it.
Based on a diff from schwarze@

rcsid[] and sccsid[] and copyright[] are essentially unmaintained (and
unmaintainable). these days, people use source. these id's do not provide
any benefit, and do hurt the small install media
(the 33,000 line diff is essentially mechanical)
ok with the idea millert, ok dms

Allow -maxdepth 0; noticed by bluhm@

Don't stop traversing a directory hierarchy if we reach SHRT_MAX,
just stop updating fts_level so we don't overflow it. This allows
rm, find, etc to operate on very deep hierarchies. Consumers of
fts(3) do need to be aware that the actual level may be larger
than fts_level. During the next libc major bump we will make
fts_level an int instead of a short. OK deraadt@

no more whiteouts

ansi; khalek@linuxgamers.net

setmode(3) returns void *, not mode_t *. ok millert@

bring protypes into scope. this requires some quirky handling, but in
the end everything is much clearer; ok tedu (itojun might like to see
how ifconfig looks after this)

mostly ansi cleanup; pval ok

Remove the advertising clause in the UCB license which Berkeley
rescinded 22 July 1999. Proofed by myself and Theo.

remove local declaration of user_from_uid()/group_from_uid()
ok millert@

Fix pasto in -anewer and -cnewer options that caused them to behave
as -newer; Dave Horsfall

Manual cleanup of remaining userland __P use (excluding packages maintained outside the tree)

Part one of userland __P removal. Done with a simple regexp with some minor hand editing to make comments line up correctly. Another pass is forthcoming that handles the cases that could not be done automatically.

kill more registers

millert@ ok

new strtofflags/fflagstostr

Add a -flags option similar to -perm but for filke flags (ala
chflags). Adapted from FreeBSD.

-iname support; karls@inet.no

o use nanosecond precision with the -newer option
o add -anewer and -cnewer options like GNU find.

If we can't chdir to the root dir in -execdir, just return instead
of generating an error. Most times this is just because the target
is not a directory. Also close an fd leak introduced in the last
commit.

Move the chdir/fchdir hack so that it is local to the -execdir
function, since that's the only place it needs to be. Doing it for
each function caused problems with multiple directories.

Do not use FTS_CHDIRROOT flag as it is fatally flawed. Instead,
do the chdir ourselves. This fixes a problem with the -execdir option
with multiple relative directories.

Add -W option so -type W gets a chance to work.

more setmode() leaks -- kill 'em all

Add the primaries -mmin, -amin, -cmin to find, similar to the GNU find; wosch

Fix problem with ``find -execdir'' not having the correct initial cwd.
Adds a new flag to fts(3).

Some -Wall cleanliness.

Add -execdir support.

Add support for -empty

Add support for -mindepth while at it

Add support for -maxdepth option

rcsid

indent nice

support whiteouts; netbsd pr#2394; mike.long@analog.com

branches: 1.1.1;
Initial revision

Account for environment when computing maximum space used for args

Already done by xargs(1). Pointed out by espie@, ok millert@

find -exec +: use sysconf to find the kernel's idea of ARG_MAX

Using ARG_MAX directly doesn't fly when ARG_MAX gets bumped and the
kernel and userland are not in sync, effectively breaking find -exec +.
Use sysconf(3) as already done in xargs(1).

Spotted by sthen@, ok deraadt@ millert@ sthen@

When system calls indicate an error they return -1, not some arbitrary
value < 0. errno is only updated in this case. Change all (most?)
callers of syscalls to follow this better, and let's see if this strictness
helps us in the future.

Use uid_from_user(3) and gid_from_group(3) in utilities that
do repeated lookups. OK tb@

add -delete option which can simplify the common case of wanting to delete
lots of files without the arcane -exec or error prone xargs.
code from freebsd.
ok millert

Convert many atoi() calls to strtonum(), adding range checks and failure
handling along the way.
Reviews by Brendan MacDonell, Jeremy Devenport, florian, doug, millert

tzfile.h is an internal header that should never have been installed.
What's worse, the tzfile.h that gets installed is over 20 years old
and doesn't match the real tzfile.h in libc/time. This makes the
tree safe for /usr/include/tzfile.h removal. The TM_YEAR_BASE
define has been moved to time.h temporarily until its usage is
replaced by 1900 in the tree. Actual removal of tzfile.h is pending
a ports build. Based on a diff from deraadt@

Replace <sys/param.h> with <limits.h> and other less dirty headers where
possible. Annotate <sys/param.h> lines with their current reasons. Switch
to PATH_MAX, NGROUPS_MAX, HOST_NAME_MAX+1, LOGIN_NAME_MAX, etc. Change
MIN() and MAX() to local definitions of MINIMUM() and MAXIMUM() where
sensible to avoid pulling in the pollution. These are the files confirmed
through binary verification.
ok guenther, millert, doug (helped with the verification protocol)

use reallocarray
okay chl@, tedu@

adjust comment

handle large numbers. this fixes time_t beyond 2038, constrains the
range of i_num correctly, and now handles files > 4GB in size on 32-bit
machines.
ok otto millert

Fix typos in comment (s,exacty,exactly,)

Add support for "find ... -exec foo {} \+".

From the manpage:

If terminated by a plus sign (``+''), the pathnames for which the
primary is evaluated are aggregated into sets, and utility will
be invoked once per set, similar to xargs(1). If any invocation
exits with non-zero exit status, then find will eventually do so
as well, but this does not cause find to exit early. The string
``{}'' must appear, and must appear last. Each set is limited to
no more than 5,000 pathnames, and is also limited such that the
invocation of utility does not exceed ARG_MAX.

Code comes from NetBSD, written by John Hawkinson. Tested by eivinde at
terraplane dot org and myself for quite some time.

ok miod@

Don't error out if we can't opendir() when -empty is used.
Fix by Bakul Shah. OK deraadt@ halex@

We only need to fchdir(dotfd) for the -exec and -ok primaries so
defer the check for dotfd != -1 until we know we will need it.
Based on a diff from schwarze@

rcsid[] and sccsid[] and copyright[] are essentially unmaintained (and
unmaintainable). these days, people use source. these id's do not provide
any benefit, and do hurt the small install media
(the 33,000 line diff is essentially mechanical)
ok with the idea millert, ok dms

Allow -maxdepth 0; noticed by bluhm@

Don't stop traversing a directory hierarchy if we reach SHRT_MAX,
just stop updating fts_level so we don't overflow it. This allows
rm, find, etc to operate on very deep hierarchies. Consumers of
fts(3) do need to be aware that the actual level may be larger
than fts_level. During the next libc major bump we will make
fts_level an int instead of a short. OK deraadt@

no more whiteouts

ansi; khalek@linuxgamers.net

setmode(3) returns void *, not mode_t *. ok millert@

bring protypes into scope. this requires some quirky handling, but in
the end everything is much clearer; ok tedu (itojun might like to see
how ifconfig looks after this)

mostly ansi cleanup; pval ok

Remove the advertising clause in the UCB license which Berkeley
rescinded 22 July 1999. Proofed by myself and Theo.

remove local declaration of user_from_uid()/group_from_uid()
ok millert@

Fix pasto in -anewer and -cnewer options that caused them to behave
as -newer; Dave Horsfall

Manual cleanup of remaining userland __P use (excluding packages maintained outside the tree)

Part one of userland __P removal. Done with a simple regexp with some minor hand editing to make comments line up correctly. Another pass is forthcoming that handles the cases that could not be done automatically.

kill more registers

millert@ ok

new strtofflags/fflagstostr

Add a -flags option similar to -perm but for filke flags (ala
chflags). Adapted from FreeBSD.

-iname support; karls@inet.no

o use nanosecond precision with the -newer option
o add -anewer and -cnewer options like GNU find.

If we can't chdir to the root dir in -execdir, just return instead
of generating an error. Most times this is just because the target
is not a directory. Also close an fd leak introduced in the last
commit.

Move the chdir/fchdir hack so that it is local to the -execdir
function, since that's the only place it needs to be. Doing it for
each function caused problems with multiple directories.

Do not use FTS_CHDIRROOT flag as it is fatally flawed. Instead,
do the chdir ourselves. This fixes a problem with the -execdir option
with multiple relative directories.

Add -W option so -type W gets a chance to work.

more setmode() leaks -- kill 'em all

Add the primaries -mmin, -amin, -cmin to find, similar to the GNU find; wosch

Fix problem with ``find -execdir'' not having the correct initial cwd.
Adds a new flag to fts(3).

Some -Wall cleanliness.

Add -execdir support.

Add support for -empty

Add support for -mindepth while at it

Add support for -maxdepth option

rcsid

indent nice

support whiteouts; netbsd pr#2394; mike.long@analog.com

branches: 1.1.1;
Initial revision

find -exec +: use sysconf to find the kernel's idea of ARG_MAX

Using ARG_MAX directly doesn't fly when ARG_MAX gets bumped and the
kernel and userland are not in sync, effectively breaking find -exec +.
Use sysconf(3) as already done in xargs(1).

Spotted by sthen@, ok deraadt@ millert@ sthen@

When system calls indicate an error they return -1, not some arbitrary
value < 0. errno is only updated in this case. Change all (most?)
callers of syscalls to follow this better, and let's see if this strictness
helps us in the future.

Use uid_from_user(3) and gid_from_group(3) in utilities that
do repeated lookups. OK tb@

add -delete option which can simplify the common case of wanting to delete
lots of files without the arcane -exec or error prone xargs.
code from freebsd.
ok millert

Convert many atoi() calls to strtonum(), adding range checks and failure
handling along the way.
Reviews by Brendan MacDonell, Jeremy Devenport, florian, doug, millert

tzfile.h is an internal header that should never have been installed.
What's worse, the tzfile.h that gets installed is over 20 years old
and doesn't match the real tzfile.h in libc/time. This makes the
tree safe for /usr/include/tzfile.h removal. The TM_YEAR_BASE
define has been moved to time.h temporarily until its usage is
replaced by 1900 in the tree. Actual removal of tzfile.h is pending
a ports build. Based on a diff from deraadt@

Replace <sys/param.h> with <limits.h> and other less dirty headers where
possible. Annotate <sys/param.h> lines with their current reasons. Switch
to PATH_MAX, NGROUPS_MAX, HOST_NAME_MAX+1, LOGIN_NAME_MAX, etc. Change
MIN() and MAX() to local definitions of MINIMUM() and MAXIMUM() where
sensible to avoid pulling in the pollution. These are the files confirmed
through binary verification.
ok guenther, millert, doug (helped with the verification protocol)

use reallocarray
okay chl@, tedu@

adjust comment

handle large numbers. this fixes time_t beyond 2038, constrains the
range of i_num correctly, and now handles files > 4GB in size on 32-bit
machines.
ok otto millert

Fix typos in comment (s,exacty,exactly,)

Add support for "find ... -exec foo {} \+".

From the manpage:

If terminated by a plus sign (``+''), the pathnames for which the
primary is evaluated are aggregated into sets, and utility will
be invoked once per set, similar to xargs(1). If any invocation
exits with non-zero exit status, then find will eventually do so
as well, but this does not cause find to exit early. The string
``{}'' must appear, and must appear last. Each set is limited to
no more than 5,000 pathnames, and is also limited such that the
invocation of utility does not exceed ARG_MAX.

Code comes from NetBSD, written by John Hawkinson. Tested by eivinde at
terraplane dot org and myself for quite some time.

ok miod@

Don't error out if we can't opendir() when -empty is used.
Fix by Bakul Shah. OK deraadt@ halex@

We only need to fchdir(dotfd) for the -exec and -ok primaries so
defer the check for dotfd != -1 until we know we will need it.
Based on a diff from schwarze@

rcsid[] and sccsid[] and copyright[] are essentially unmaintained (and
unmaintainable). these days, people use source. these id's do not provide
any benefit, and do hurt the small install media
(the 33,000 line diff is essentially mechanical)
ok with the idea millert, ok dms

Allow -maxdepth 0; noticed by bluhm@

Don't stop traversing a directory hierarchy if we reach SHRT_MAX,
just stop updating fts_level so we don't overflow it. This allows
rm, find, etc to operate on very deep hierarchies. Consumers of
fts(3) do need to be aware that the actual level may be larger
than fts_level. During the next libc major bump we will make
fts_level an int instead of a short. OK deraadt@

no more whiteouts

ansi; khalek@linuxgamers.net

setmode(3) returns void *, not mode_t *. ok millert@

bring protypes into scope. this requires some quirky handling, but in
the end everything is much clearer; ok tedu (itojun might like to see
how ifconfig looks after this)

mostly ansi cleanup; pval ok

Remove the advertising clause in the UCB license which Berkeley
rescinded 22 July 1999. Proofed by myself and Theo.

remove local declaration of user_from_uid()/group_from_uid()
ok millert@

Fix pasto in -anewer and -cnewer options that caused them to behave
as -newer; Dave Horsfall

Manual cleanup of remaining userland __P use (excluding packages maintained outside the tree)

Part one of userland __P removal. Done with a simple regexp with some minor hand editing to make comments line up correctly. Another pass is forthcoming that handles the cases that could not be done automatically.

kill more registers

millert@ ok

new strtofflags/fflagstostr

Add a -flags option similar to -perm but for filke flags (ala
chflags). Adapted from FreeBSD.

-iname support; karls@inet.no

o use nanosecond precision with the -newer option
o add -anewer and -cnewer options like GNU find.

If we can't chdir to the root dir in -execdir, just return instead
of generating an error. Most times this is just because the target
is not a directory. Also close an fd leak introduced in the last
commit.

Move the chdir/fchdir hack so that it is local to the -execdir
function, since that's the only place it needs to be. Doing it for
each function caused problems with multiple directories.

Do not use FTS_CHDIRROOT flag as it is fatally flawed. Instead,
do the chdir ourselves. This fixes a problem with the -execdir option
with multiple relative directories.

Add -W option so -type W gets a chance to work.

more setmode() leaks -- kill 'em all

Add the primaries -mmin, -amin, -cmin to find, similar to the GNU find; wosch

Fix problem with ``find -execdir'' not having the correct initial cwd.
Adds a new flag to fts(3).

Some -Wall cleanliness.

Add -execdir support.

Add support for -empty

Add support for -mindepth while at it

Add support for -maxdepth option

rcsid

indent nice

support whiteouts; netbsd pr#2394; mike.long@analog.com

branches: 1.1.1;
Initial revision

When system calls indicate an error they return -1, not some arbitrary
value < 0. errno is only updated in this case. Change all (most?)
callers of syscalls to follow this better, and let's see if this strictness
helps us in the future.

Use uid_from_user(3) and gid_from_group(3) in utilities that
do repeated lookups. OK tb@

add -delete option which can simplify the common case of wanting to delete
lots of files without the arcane -exec or error prone xargs.
code from freebsd.
ok millert

Convert many atoi() calls to strtonum(), adding range checks and failure
handling along the way.
Reviews by Brendan MacDonell, Jeremy Devenport, florian, doug, millert

tzfile.h is an internal header that should never have been installed.
What's worse, the tzfile.h that gets installed is over 20 years old
and doesn't match the real tzfile.h in libc/time. This makes the
tree safe for /usr/include/tzfile.h removal. The TM_YEAR_BASE
define has been moved to time.h temporarily until its usage is
replaced by 1900 in the tree. Actual removal of tzfile.h is pending
a ports build. Based on a diff from deraadt@

Replace <sys/param.h> with <limits.h> and other less dirty headers where
possible. Annotate <sys/param.h> lines with their current reasons. Switch
to PATH_MAX, NGROUPS_MAX, HOST_NAME_MAX+1, LOGIN_NAME_MAX, etc. Change
MIN() and MAX() to local definitions of MINIMUM() and MAXIMUM() where
sensible to avoid pulling in the pollution. These are the files confirmed
through binary verification.
ok guenther, millert, doug (helped with the verification protocol)

use reallocarray
okay chl@, tedu@

adjust comment

handle large numbers. this fixes time_t beyond 2038, constrains the
range of i_num correctly, and now handles files > 4GB in size on 32-bit
machines.
ok otto millert

Fix typos in comment (s,exacty,exactly,)

Add support for "find ... -exec foo {} \+".

From the manpage:

If terminated by a plus sign (``+''), the pathnames for which the
primary is evaluated are aggregated into sets, and utility will
be invoked once per set, similar to xargs(1). If any invocation
exits with non-zero exit status, then find will eventually do so
as well, but this does not cause find to exit early. The string
``{}'' must appear, and must appear last. Each set is limited to
no more than 5,000 pathnames, and is also limited such that the
invocation of utility does not exceed ARG_MAX.

Code comes from NetBSD, written by John Hawkinson. Tested by eivinde at
terraplane dot org and myself for quite some time.

ok miod@

Don't error out if we can't opendir() when -empty is used.
Fix by Bakul Shah. OK deraadt@ halex@

We only need to fchdir(dotfd) for the -exec and -ok primaries so
defer the check for dotfd != -1 until we know we will need it.
Based on a diff from schwarze@

rcsid[] and sccsid[] and copyright[] are essentially unmaintained (and
unmaintainable). these days, people use source. these id's do not provide
any benefit, and do hurt the small install media
(the 33,000 line diff is essentially mechanical)
ok with the idea millert, ok dms

Allow -maxdepth 0; noticed by bluhm@

Don't stop traversing a directory hierarchy if we reach SHRT_MAX,
just stop updating fts_level so we don't overflow it. This allows
rm, find, etc to operate on very deep hierarchies. Consumers of
fts(3) do need to be aware that the actual level may be larger
than fts_level. During the next libc major bump we will make
fts_level an int instead of a short. OK deraadt@

no more whiteouts

ansi; khalek@linuxgamers.net

setmode(3) returns void *, not mode_t *. ok millert@

bring protypes into scope. this requires some quirky handling, but in
the end everything is much clearer; ok tedu (itojun might like to see
how ifconfig looks after this)

mostly ansi cleanup; pval ok

Remove the advertising clause in the UCB license which Berkeley
rescinded 22 July 1999. Proofed by myself and Theo.

remove local declaration of user_from_uid()/group_from_uid()
ok millert@

Fix pasto in -anewer and -cnewer options that caused them to behave
as -newer; Dave Horsfall

Manual cleanup of remaining userland __P use (excluding packages maintained outside the tree)

Part one of userland __P removal. Done with a simple regexp with some minor hand editing to make comments line up correctly. Another pass is forthcoming that handles the cases that could not be done automatically.

kill more registers

millert@ ok

new strtofflags/fflagstostr

Add a -flags option similar to -perm but for filke flags (ala
chflags). Adapted from FreeBSD.

-iname support; karls@inet.no

o use nanosecond precision with the -newer option
o add -anewer and -cnewer options like GNU find.

If we can't chdir to the root dir in -execdir, just return instead
of generating an error. Most times this is just because the target
is not a directory. Also close an fd leak introduced in the last
commit.

Move the chdir/fchdir hack so that it is local to the -execdir
function, since that's the only place it needs to be. Doing it for
each function caused problems with multiple directories.

Do not use FTS_CHDIRROOT flag as it is fatally flawed. Instead,
do the chdir ourselves. This fixes a problem with the -execdir option
with multiple relative directories.

Add -W option so -type W gets a chance to work.

more setmode() leaks -- kill 'em all

Add the primaries -mmin, -amin, -cmin to find, similar to the GNU find; wosch

Fix problem with ``find -execdir'' not having the correct initial cwd.
Adds a new flag to fts(3).

Some -Wall cleanliness.

Add -execdir support.

Add support for -empty

Add support for -mindepth while at it

Add support for -maxdepth option

rcsid

indent nice

support whiteouts; netbsd pr#2394; mike.long@analog.com

branches: 1.1.1;
Initial revision

Use uid_from_user(3) and gid_from_group(3) in utilities that
do repeated lookups. OK tb@

add -delete option which can simplify the common case of wanting to delete
lots of files without the arcane -exec or error prone xargs.
code from freebsd.
ok millert

Convert many atoi() calls to strtonum(), adding range checks and failure
handling along the way.
Reviews by Brendan MacDonell, Jeremy Devenport, florian, doug, millert

tzfile.h is an internal header that should never have been installed.
What's worse, the tzfile.h that gets installed is over 20 years old
and doesn't match the real tzfile.h in libc/time. This makes the
tree safe for /usr/include/tzfile.h removal. The TM_YEAR_BASE
define has been moved to time.h temporarily until its usage is
replaced by 1900 in the tree. Actual removal of tzfile.h is pending
a ports build. Based on a diff from deraadt@

Replace <sys/param.h> with <limits.h> and other less dirty headers where
possible. Annotate <sys/param.h> lines with their current reasons. Switch
to PATH_MAX, NGROUPS_MAX, HOST_NAME_MAX+1, LOGIN_NAME_MAX, etc. Change
MIN() and MAX() to local definitions of MINIMUM() and MAXIMUM() where
sensible to avoid pulling in the pollution. These are the files confirmed
through binary verification.
ok guenther, millert, doug (helped with the verification protocol)

use reallocarray
okay chl@, tedu@

adjust comment

handle large numbers. this fixes time_t beyond 2038, constrains the
range of i_num correctly, and now handles files > 4GB in size on 32-bit
machines.
ok otto millert

Fix typos in comment (s,exacty,exactly,)

Add support for "find ... -exec foo {} \+".

From the manpage:

If terminated by a plus sign (``+''), the pathnames for which the
primary is evaluated are aggregated into sets, and utility will
be invoked once per set, similar to xargs(1). If any invocation
exits with non-zero exit status, then find will eventually do so
as well, but this does not cause find to exit early. The string
``{}'' must appear, and must appear last. Each set is limited to
no more than 5,000 pathnames, and is also limited such that the
invocation of utility does not exceed ARG_MAX.

Code comes from NetBSD, written by John Hawkinson. Tested by eivinde at
terraplane dot org and myself for quite some time.

ok miod@

Don't error out if we can't opendir() when -empty is used.
Fix by Bakul Shah. OK deraadt@ halex@

We only need to fchdir(dotfd) for the -exec and -ok primaries so
defer the check for dotfd != -1 until we know we will need it.
Based on a diff from schwarze@

rcsid[] and sccsid[] and copyright[] are essentially unmaintained (and
unmaintainable). these days, people use source. these id's do not provide
any benefit, and do hurt the small install media
(the 33,000 line diff is essentially mechanical)
ok with the idea millert, ok dms

Allow -maxdepth 0; noticed by bluhm@

Don't stop traversing a directory hierarchy if we reach SHRT_MAX,
just stop updating fts_level so we don't overflow it. This allows
rm, find, etc to operate on very deep hierarchies. Consumers of
fts(3) do need to be aware that the actual level may be larger
than fts_level. During the next libc major bump we will make
fts_level an int instead of a short. OK deraadt@

no more whiteouts

ansi; khalek@linuxgamers.net