moncontrol(3): remove hertz() fallback function

In the moncontrol(3) code in libc there is a fallback function,
hertz(). The idea is, if getting kern.clockrate from sysctl(2) fails,
we fall back to deriving the value of hz(9) using setitimer(2)'s
rounding behavior.

This is extremely clever, but it actually sucks. Calling setitimer(2)
quietly cancels any extant ITIMER_REAL timer, so moncontrol(3) cannot
be safely used alongside setitimer(2). This fact is not documented.

kern.clockrate is not blocked by pledge(2), so outside of stack
corruption (which we can't do anything about anyway) I don't see a way
for the sysctl(2) call to ever fail on OpenBSD. So hertz() is also
pointless.

Hence this patch: get rid of hertz().

Thread: https://marc.info/?l=openbsd-tech&m=163881542813633&w=2

ok guenther@

make fixed-sized fixed-value mib[] arrays be const
ok guenther tb millert

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.

Delete casts to off_t and size_t that are implied by assignments
or prototypes. Ditto for some of the char* and void* casts too.

verified no change to instructions on ILP32 (i386) and LP64 (amd64)
ok natano@ abluhm@ deraadt@ millert@

Declare moncontrol(3) APIs in <sys/gmon.h>
Export _gmonparam again.
Make gcrt0.o use an reserved name for _monstartup()

ok millert@

(char *)0 -> NULL

environ and __progname are not declared in a public header; declare them
in libc's hidden/stdlib.h instead of in each .c file that needs one

ok deraadt@ gsoares@ mpi@

replace (void *)0 with NULL

Wrap moncontrol() so that internal calls go direct and it's a weak symbol

fix a fd leak in an error path in code under #ifdef DEBUG

Move to the <limits.h> universe.
review by millert, binary checking process with doug, concept with guenther

Back out per-CPU kernel profiling, it shouldn't modify a public header
at this moment.

Unbreak the tree by adding missing defines. Forgot to sync in previous
spotted by todd@.

Allocate space for profiling data using mmap instead of sbrk.
Tested by myself and pedro@

zap remaining rcsid.

Kill old files that are no longer compiled.

okay theo

be more careful about snprintf return value; ok beck cloder

deregister & ansify. ok deraad@ marco@ cloder@

protos

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rescinded 22 July 1999.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

Zero out the reserved fields of struct gmonhdr rather than leaving
stack garbage in them. From FreeBSD.

OK deraadt@

remaining easy snprintf conversions

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.

make that 0664 until we find out why umask is not being applied

open gmon.out mode 0644, not 0666

use STDERR_FILENO; kleink

tabify + trailing blanks

Remove dead code
Remove unused variables
Silence some warnings
lint(1) is your friend

protect getenv, do not oflow, fix unused debug; bitblt and deraadt

Fix RCS ids
Make sure everything uses {SYS,}LIBC_SCCS properly

use p->hashfraction when doing non-time-critical calculations, rather than
using HASHFRACTION directly. in time-critical calculations, if HASHFRACTION
is a power of two, check that p->hashfraction == HASHFRACTION and if so do
the calculation with the compiled-in value so that the compiler can optimize
out (potentially) expensive divisions. if p->hashfraction != HASHFRACTION,
actually do the division. This has the result that on machines with slow
division, the division can be optimized out of the common case, but that
if HASHFRACTION changes from the compiled-in value (for whatever reason),
profiling will still work. Changes suggested by Chris Torek.

Protect internal mcount symbol from lint(1)
Use __asm, not asm

from cgd@netbsd:
kill a gratuitous cast of lowpc to an int
Add support for PROFDIR environment variable

branches: 1.1.1;
Initial revision

make fixed-sized fixed-value mib[] arrays be const
ok guenther tb millert

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.

Delete casts to off_t and size_t that are implied by assignments
or prototypes. Ditto for some of the char* and void* casts too.

verified no change to instructions on ILP32 (i386) and LP64 (amd64)
ok natano@ abluhm@ deraadt@ millert@

Declare moncontrol(3) APIs in <sys/gmon.h>
Export _gmonparam again.
Make gcrt0.o use an reserved name for _monstartup()

ok millert@

(char *)0 -> NULL

environ and __progname are not declared in a public header; declare them
in libc's hidden/stdlib.h instead of in each .c file that needs one

ok deraadt@ gsoares@ mpi@

replace (void *)0 with NULL

Wrap moncontrol() so that internal calls go direct and it's a weak symbol

fix a fd leak in an error path in code under #ifdef DEBUG

Move to the <limits.h> universe.
review by millert, binary checking process with doug, concept with guenther

Back out per-CPU kernel profiling, it shouldn't modify a public header
at this moment.

Unbreak the tree by adding missing defines. Forgot to sync in previous
spotted by todd@.

Allocate space for profiling data using mmap instead of sbrk.
Tested by myself and pedro@

zap remaining rcsid.

Kill old files that are no longer compiled.

okay theo

be more careful about snprintf return value; ok beck cloder

deregister & ansify. ok deraad@ marco@ cloder@

protos

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

Zero out the reserved fields of struct gmonhdr rather than leaving
stack garbage in them. From FreeBSD.

OK deraadt@

remaining easy snprintf conversions

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.

make that 0664 until we find out why umask is not being applied

open gmon.out mode 0644, not 0666

use STDERR_FILENO; kleink

tabify + trailing blanks

Remove dead code
Remove unused variables
Silence some warnings
lint(1) is your friend

protect getenv, do not oflow, fix unused debug; bitblt and deraadt

Fix RCS ids
Make sure everything uses {SYS,}LIBC_SCCS properly

use p->hashfraction when doing non-time-critical calculations, rather than
using HASHFRACTION directly. in time-critical calculations, if HASHFRACTION
is a power of two, check that p->hashfraction == HASHFRACTION and if so do
the calculation with the compiled-in value so that the compiler can optimize
out (potentially) expensive divisions. if p->hashfraction != HASHFRACTION,
actually do the division. This has the result that on machines with slow
division, the division can be optimized out of the common case, but that
if HASHFRACTION changes from the compiled-in value (for whatever reason),
profiling will still work. Changes suggested by Chris Torek.

Protect internal mcount symbol from lint(1)
Use __asm, not asm

from cgd@netbsd:
kill a gratuitous cast of lowpc to an int
Add support for PROFDIR environment variable

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.

Delete casts to off_t and size_t that are implied by assignments
or prototypes. Ditto for some of the char* and void* casts too.

verified no change to instructions on ILP32 (i386) and LP64 (amd64)
ok natano@ abluhm@ deraadt@ millert@

Declare moncontrol(3) APIs in <sys/gmon.h>
Export _gmonparam again.
Make gcrt0.o use an reserved name for _monstartup()

ok millert@

(char *)0 -> NULL

environ and __progname are not declared in a public header; declare them
in libc's hidden/stdlib.h instead of in each .c file that needs one

ok deraadt@ gsoares@ mpi@

replace (void *)0 with NULL

Wrap moncontrol() so that internal calls go direct and it's a weak symbol

fix a fd leak in an error path in code under #ifdef DEBUG

Move to the <limits.h> universe.
review by millert, binary checking process with doug, concept with guenther

Back out per-CPU kernel profiling, it shouldn't modify a public header
at this moment.

Unbreak the tree by adding missing defines. Forgot to sync in previous
spotted by todd@.

Allocate space for profiling data using mmap instead of sbrk.
Tested by myself and pedro@

zap remaining rcsid.

Kill old files that are no longer compiled.

okay theo

be more careful about snprintf return value; ok beck cloder

deregister & ansify. ok deraad@ marco@ cloder@

protos

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

Zero out the reserved fields of struct gmonhdr rather than leaving
stack garbage in them. From FreeBSD.

OK deraadt@

remaining easy snprintf conversions

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.

make that 0664 until we find out why umask is not being applied

open gmon.out mode 0644, not 0666

use STDERR_FILENO; kleink

tabify + trailing blanks

Remove dead code
Remove unused variables
Silence some warnings
lint(1) is your friend

protect getenv, do not oflow, fix unused debug; bitblt and deraadt

Fix RCS ids
Make sure everything uses {SYS,}LIBC_SCCS properly

use p->hashfraction when doing non-time-critical calculations, rather than
using HASHFRACTION directly. in time-critical calculations, if HASHFRACTION
is a power of two, check that p->hashfraction == HASHFRACTION and if so do
the calculation with the compiled-in value so that the compiler can optimize
out (potentially) expensive divisions. if p->hashfraction != HASHFRACTION,
actually do the division. This has the result that on machines with slow
division, the division can be optimized out of the common case, but that
if HASHFRACTION changes from the compiled-in value (for whatever reason),
profiling will still work. Changes suggested by Chris Torek.

Protect internal mcount symbol from lint(1)
Use __asm, not asm

from cgd@netbsd:
kill a gratuitous cast of lowpc to an int
Add support for PROFDIR environment variable

branches: 1.1.1;
Initial revision

Delete casts to off_t and size_t that are implied by assignments
or prototypes. Ditto for some of the char* and void* casts too.

verified no change to instructions on ILP32 (i386) and LP64 (amd64)
ok natano@ abluhm@ deraadt@ millert@

Declare moncontrol(3) APIs in <sys/gmon.h>
Export _gmonparam again.
Make gcrt0.o use an reserved name for _monstartup()

ok millert@

(char *)0 -> NULL

environ and __progname are not declared in a public header; declare them
in libc's hidden/stdlib.h instead of in each .c file that needs one

ok deraadt@ gsoares@ mpi@

replace (void *)0 with NULL

Wrap moncontrol() so that internal calls go direct and it's a weak symbol

fix a fd leak in an error path in code under #ifdef DEBUG

Move to the <limits.h> universe.
review by millert, binary checking process with doug, concept with guenther

Back out per-CPU kernel profiling, it shouldn't modify a public header
at this moment.

Unbreak the tree by adding missing defines. Forgot to sync in previous
spotted by todd@.

Allocate space for profiling data using mmap instead of sbrk.
Tested by myself and pedro@

zap remaining rcsid.

Kill old files that are no longer compiled.

okay theo

be more careful about snprintf return value; ok beck cloder

deregister & ansify. ok deraad@ marco@ cloder@

protos

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

Zero out the reserved fields of struct gmonhdr rather than leaving
stack garbage in them. From FreeBSD.

OK deraadt@

remaining easy snprintf conversions

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.

make that 0664 until we find out why umask is not being applied

open gmon.out mode 0644, not 0666

use STDERR_FILENO; kleink

tabify + trailing blanks

Remove dead code
Remove unused variables
Silence some warnings
lint(1) is your friend

protect getenv, do not oflow, fix unused debug; bitblt and deraadt

Fix RCS ids
Make sure everything uses {SYS,}LIBC_SCCS properly

use p->hashfraction when doing non-time-critical calculations, rather than
using HASHFRACTION directly. in time-critical calculations, if HASHFRACTION
is a power of two, check that p->hashfraction == HASHFRACTION and if so do
the calculation with the compiled-in value so that the compiler can optimize
out (potentially) expensive divisions. if p->hashfraction != HASHFRACTION,
actually do the division. This has the result that on machines with slow
division, the division can be optimized out of the common case, but that
if HASHFRACTION changes from the compiled-in value (for whatever reason),
profiling will still work. Changes suggested by Chris Torek.

Protect internal mcount symbol from lint(1)
Use __asm, not asm

from cgd@netbsd:
kill a gratuitous cast of lowpc to an int
Add support for PROFDIR environment variable

branches: 1.1.1;
Initial revision