mischief/openbsd-src
1
0
Fork 0
mirror of https://github.com/openbsd/src.git synced 2024-12-22 16:42:56 -08:00
Code
master
openbsd-src/bin/dd/misc.c

110 lines
3.4 KiB
C
Raw Permalink Normal View History

Rewrite bytes/sec calculation using fixed point math. This makes signal handler safe on OpenBSD. To avoid overflows the accuracy is scaled. Above 10 minutes run time we only care about second accuracy. Between 1 seconds and 10 minutes we use millisecond accuracy. Below one second we use nanoseconds, but those numbers are probably meaningless. Signal handler problem pointed out by deraadt OK deraadt, millert, tb
2024-07-12 12:11:25 -07:00
/* $OpenBSD: misc.c,v 1.26 2024/07/12 19:11:25 florian Exp $ */
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
/* $NetBSD: misc.c,v 1.4 1995/03/21 09:04:10 cgd Exp $ */
/*-
* Copyright (c) 1991, 1993, 1994
* The Regents of the University of California. All rights reserved.
*
* This code is derived from software contributed to Berkeley by
* Keith Muller of the University of California, San Diego and Lance
* Visser of Convex Computer Corporation.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
Remove the advertising clause in the UCB license which Berkeley rescinded 22 July 1999. Proofed by myself and Theo.
2003-06-02 16:27:43 -07:00
* 3. Neither the name of the University nor the names of its contributors
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/types.h>
include sys/time.h to get struct timeval
2005-05-26 21:14:24 -07:00
#include <sys/time.h>
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
errno.h
1997-08-05 15:50:51 -07:00
#include <errno.h>
snprintf+writev -> dprintf Much simpler. Dropping writev makes the output from summary() non-atomic, so output lines triggered by SIGINFO can intermingle with output lines triggered by SIGINT or atexit(3). This behavior is difficult to trigger, even deliberately. The buffer in dprintf is large enough so that lines produced by summary() are not garbled if this occurs, though, so this is unlikely to break any scripts. Discussed at length with deraadt@, who helpfully noted that "not all problems have perfect solutions." ok millert@
2018-04-07 11:52:39 -07:00
#include <stdio.h>
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
#include <time.h>
#include <unistd.h>
#include "dd.h"
#include "extern.h"
refactor the signal handlers for clarity, inverting the situation: the signal handler was calling a big function which is shared between multiple contexts -- that hides the rule that this big function has signal safe requirements (which it fails). now, the signal handler contains all the code, and everyone else calls the signal handler function as a regular function, from their (normal) contexts. the signal handler context is the most strict, so this pattern is better. ok florian
2024-07-12 07:30:27 -07:00
/* SIGINFO handler */
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
void
refactor the signal handlers for clarity, inverting the situation: the signal handler was calling a big function which is shared between multiple contexts -- that hides the rule that this big function has signal safe requirements (which it fails). now, the signal handler contains all the code, and everyone else calls the signal handler function as a regular function, from their (normal) contexts. the signal handler context is the most strict, so this pattern is better. ok florian
2024-07-12 07:30:27 -07:00
sig_summary(int notused)
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
{
refactor the signal handlers for clarity, inverting the situation: the signal handler was calling a big function which is shared between multiple contexts -- that hides the rule that this big function has signal safe requirements (which it fails). now, the signal handler contains all the code, and everyone else calls the signal handler function as a regular function, from their (normal) contexts. the signal handler context is the most strict, so this pattern is better. ok florian
2024-07-12 07:30:27 -07:00
int save_errno = errno;
convert gettimeofday to mono clock. from Scott Cheloha
2017-08-12 19:06:42 -07:00
struct timespec elapsed, now;
Rewrite bytes/sec calculation using fixed point math. This makes signal handler safe on OpenBSD. To avoid overflows the accuracy is scaled. Above 10 minutes run time we only care about second accuracy. Between 1 seconds and 10 minutes we use millisecond accuracy. Below one second we use nanoseconds, but those numbers are probably meaningless. Signal handler problem pointed out by deraadt OK deraadt, millert, tb
2024-07-12 12:11:25 -07:00
unsigned long long bps, msec;
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
Add an operand to /bin/dd that suppresses the status lines printed to stderr after transfer. Use the GNU syntax status=noxfer and status=none to make dd quiet. The output is annoying in some situations, so people redirect stderr to /dev/null. This approach also suppresses the error messages and bugs in shell scripts remain undetected. The new feature status=none is the better alternative. OK mpf@
2014-02-11 17:18:36 -08:00
if (ddflags & C_NOINFO)
return;
convert gettimeofday to mono clock. from Scott Cheloha
2017-08-12 19:06:42 -07:00
clock_gettime(CLOCK_MONOTONIC, &now);
timespecsub(&now, &st.start, &elapsed);
Rewrite bytes/sec calculation using fixed point math. This makes signal handler safe on OpenBSD. To avoid overflows the accuracy is scaled. Above 10 minutes run time we only care about second accuracy. Between 1 seconds and 10 minutes we use millisecond accuracy. Below one second we use nanoseconds, but those numbers are probably meaningless. Signal handler problem pointed out by deraadt OK deraadt, millert, tb
2024-07-12 12:11:25 -07:00
if (elapsed.tv_sec > 600)
bps = st.bytes / elapsed.tv_sec;
else if (elapsed.tv_sec > 0) {
/* will overflow at ~ 30 exabytes / second */
msec = elapsed.tv_sec * 1000 + elapsed.tv_nsec / 1000000;
if (msec == 0)
msec = 1;
bps = st.bytes * 1000 / msec;
} else if (elapsed.tv_nsec > 0)
bps = st.bytes * 1000000000 / elapsed.tv_nsec;
else
bps = st.bytes;
Modify timekeeping in an effort to produce a sane bandwidth measure for a short running dd process. Some floating point is used so that microsecond precision can be kept without shortening the off_t excessively. millert says ok.
2001-08-07 07:39:27 -07:00
snprintf+writev -> dprintf Much simpler. Dropping writev makes the output from summary() non-atomic, so output lines triggered by SIGINFO can intermingle with output lines triggered by SIGINT or atexit(3). This behavior is difficult to trigger, even deliberately. The buffer in dprintf is large enough so that lines produced by summary() are not garbled if this occurs, though, so this is unlikely to break any scripts. Discussed at length with deraadt@, who helpfully noted that "not all problems have perfect solutions." ok millert@
2018-04-07 11:52:39 -07:00
/* Be async safe: use dprintf(3). */
dprintf(STDERR_FILENO, "%zu+%zu records in\n%zu+%zu records out\n",
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
st.in_full, st.in_part, st.out_full, st.out_part);
had not noticed the atexit() handler; pointed out by dima@unixfreak.org; also use writev to increase atomicity
2001-06-25 01:06:04 -07:00
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
if (st.swab) {
snprintf+writev -> dprintf Much simpler. Dropping writev makes the output from summary() non-atomic, so output lines triggered by SIGINFO can intermingle with output lines triggered by SIGINT or atexit(3). This behavior is difficult to trigger, even deliberately. The buffer in dprintf is large enough so that lines produced by summary() are not garbled if this occurs, though, so this is unlikely to break any scripts. Discussed at length with deraadt@, who helpfully noted that "not all problems have perfect solutions." ok millert@
2018-04-07 11:52:39 -07:00
dprintf(STDERR_FILENO, "%zu odd length swab %s\n",
st.swab, (st.swab == 1) ? "block" : "blocks");
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
}
if (st.trunc) {
snprintf+writev -> dprintf Much simpler. Dropping writev makes the output from summary() non-atomic, so output lines triggered by SIGINFO can intermingle with output lines triggered by SIGINT or atexit(3). This behavior is difficult to trigger, even deliberately. The buffer in dprintf is large enough so that lines produced by summary() are not garbled if this occurs, though, so this is unlikely to break any scripts. Discussed at length with deraadt@, who helpfully noted that "not all problems have perfect solutions." ok millert@
2018-04-07 11:52:39 -07:00
dprintf(STDERR_FILENO, "%zu truncated %s\n",
st.trunc, (st.trunc == 1) ? "block" : "blocks");
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
}
Add an operand to /bin/dd that suppresses the status lines printed to stderr after transfer. Use the GNU syntax status=noxfer and status=none to make dd quiet. The output is annoying in some situations, so people redirect stderr to /dev/null. This approach also suppresses the error messages and bugs in shell scripts remain undetected. The new feature status=none is the better alternative. OK mpf@
2014-02-11 17:18:36 -08:00
if (!(ddflags & C_NOXFER)) {
snprintf+writev -> dprintf Much simpler. Dropping writev makes the output from summary() non-atomic, so output lines triggered by SIGINFO can intermingle with output lines triggered by SIGINT or atexit(3). This behavior is difficult to trigger, even deliberately. The buffer in dprintf is large enough so that lines produced by summary() are not garbled if this occurs, though, so this is unlikely to break any scripts. Discussed at length with deraadt@, who helpfully noted that "not all problems have perfect solutions." ok millert@
2018-04-07 11:52:39 -07:00
dprintf(STDERR_FILENO,
%qd -> %lld for an off_t already being cast to (long long)!
2016-08-14 14:13:08 -07:00
"%lld bytes transferred in %lld.%03ld secs "
Rewrite bytes/sec calculation using fixed point math. This makes signal handler safe on OpenBSD. To avoid overflows the accuracy is scaled. Above 10 minutes run time we only care about second accuracy. Between 1 seconds and 10 minutes we use millisecond accuracy. Below one second we use nanoseconds, but those numbers are probably meaningless. Signal handler problem pointed out by deraadt OK deraadt, millert, tb
2024-07-12 12:11:25 -07:00
"(%llu bytes/sec)\n", (long long)st.bytes,
(long long)elapsed.tv_sec, elapsed.tv_nsec / 1000000, bps);
Add an operand to /bin/dd that suppresses the status lines printed to stderr after transfer. Use the GNU syntax status=noxfer and status=none to make dd quiet. The output is annoying in some situations, so people redirect stderr to /dev/null. This approach also suppresses the error messages and bugs in shell scripts remain undetected. The new feature status=none is the better alternative. OK mpf@
2014-02-11 17:18:36 -08:00
}
refactor the signal handlers for clarity, inverting the situation: the signal handler was calling a big function which is shared between multiple contexts -- that hides the rule that this big function has signal safe requirements (which it fails). now, the signal handler contains all the code, and everyone else calls the signal handler function as a regular function, from their (normal) contexts. the signal handler context is the most strict, so this pattern is better. ok florian
2024-07-12 07:30:27 -07:00
errno = save_errno;
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
}
refactor the signal handlers for clarity, inverting the situation: the signal handler was calling a big function which is shared between multiple contexts -- that hides the rule that this big function has signal safe requirements (which it fails). now, the signal handler contains all the code, and everyone else calls the signal handler function as a regular function, from their (normal) contexts. the signal handler context is the most strict, so this pattern is better. ok florian
2024-07-12 07:30:27 -07:00
/* SIGINT handler */
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
void
refactor the signal handlers for clarity, inverting the situation: the signal handler was calling a big function which is shared between multiple contexts -- that hides the rule that this big function has signal safe requirements (which it fails). now, the signal handler contains all the code, and everyone else calls the signal handler function as a regular function, from their (normal) contexts. the signal handler context is the most strict, so this pattern is better. ok florian
2024-07-12 07:30:27 -07:00
sig_terminate(int signo)
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
{
refactor the signal handlers for clarity, inverting the situation: the signal handler was calling a big function which is shared between multiple contexts -- that hides the rule that this big function has signal safe requirements (which it fails). now, the signal handler contains all the code, and everyone else calls the signal handler function as a regular function, from their (normal) contexts. the signal handler context is the most strict, so this pattern is better. ok florian
2024-07-12 07:30:27 -07:00
sig_summary(0);
_exit(128 + signo);
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
}
refactor the signal handlers for clarity, inverting the situation: the signal handler was calling a big function which is shared between multiple contexts -- that hides the rule that this big function has signal safe requirements (which it fails). now, the signal handler contains all the code, and everyone else calls the signal handler function as a regular function, from their (normal) contexts. the signal handler context is the most strict, so this pattern is better. ok florian
2024-07-12 07:30:27 -07:00
/* atexit variation to summarize */
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
void
refactor the signal handlers for clarity, inverting the situation: the signal handler was calling a big function which is shared between multiple contexts -- that hides the rule that this big function has signal safe requirements (which it fails). now, the signal handler contains all the code, and everyone else calls the signal handler function as a regular function, from their (normal) contexts. the signal handler context is the most strict, so this pattern is better. ok florian
2024-07-12 07:30:27 -07:00
exit_summary(void)
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
{
refactor the signal handlers for clarity, inverting the situation: the signal handler was calling a big function which is shared between multiple contexts -- that hides the rule that this big function has signal safe requirements (which it fails). now, the signal handler contains all the code, and everyone else calls the signal handler function as a regular function, from their (normal) contexts. the signal handler context is the most strict, so this pattern is better. ok florian
2024-07-12 07:30:27 -07:00
sig_summary(0);
initial import of NetBSD tree
1995-10-18 01:37:01 -07:00
}
Copy Permalink