bad42df9 cperciva Sept. 5, 2021, 7:48 p.m.
About 335 ms of EC2 instance boot time is being spent here.
cgit
7347dfce cperciva Sept. 5, 2021, 7:45 p.m.
The 'intr_config_hooks' SYSINIT is now taking a nontrivial amount of
time in my profiling; run_interrupt_driven_config_hooks is responsible
for most of it, so this adds useful information to the resulting
flamecharts.
cgit
007724cb mjg Sept. 5, 2021, 5:31 p.m.
bb61ccd5 pho Sept. 5, 2021, 7:01 a.m.
1f7a6325 mav Sept. 4, 2021, 10:30 p.m.
None of the drivers really implement asynchronous buffer loading, so
they don't need lock there.  The only question is which of them have
BUS_DMA_NOWAIT flag there already, and to which it should be added.
cgit
440320b6 rew Sept. 4, 2021, 8:25 p.m.
MFC After:      1 week
Reviewed by:	imp, kib
Differential Revision:	https://reviews.freebsd.org/D31127
cgit
541699f6 mav Sept. 4, 2021, 8:10 p.m.
71bf3900 mav Sept. 4, 2021, 7:54 p.m.
MFC after:	1 month
cgit
312f5f8a kp Sept. 4, 2021, 3:26 p.m.
There's no reason to acquire the Giant lock while executing the ALTQ
callouts.

While here also remove a few backwards compatibility defines for long
obsolete FreeBSD versions.

Reviewed by:	mav
Suggested by:	mav
Differential Revision:	https://reviews.freebsd.org/D31835
cgit
7b81e6c7 jhibbits Sept. 4, 2021, 4:47 p.m.
eb937f67 jhibbits Sept. 4, 2021, 4:44 p.m.
Should be saving off td0's register's, not td's.
cgit
35a1aa5f jhibbits Sept. 4, 2021, 4:33 p.m.
Summary:
Failure to update the FP / vector state was causing daemon(3) to violate C ABI by failing to preserve nonvolatile registers.

This was causing a weird issue where moused was not working on PowerBook G4s when daemonizing, but was working fine when running it foreground.

Force saving off the same state that cpu_switch() does in cases where we are about to copy a thread.

MFC after: 1 week
Sponsored by: Tag1 Consulting, Inc.

Test Plan:
```
/*
 * Test for ABI violation due to side effects of daemon(3).
 *
 * NOTE: Compile with -O2 to see the effect.
 */
/* Allow compiling for Linux too. */

static double test = 1234.56f;

/*
 * This contrivance coerces clang to not bounce the double
 * off of memory again in main.
 */
void __attribute__((noinline))
print_double(int j1, int j2, double d)
{
	printf("%f\n", d);
}

int
main(int argc, char *argv[])
{
	print_double(0, 0, test);

	if (daemon(0, 1)) {
	}
	/* Compiler assumes nonvolatile regs are intact... */
	print_double(0, 0, test);
	return(0);
}
```

Working output:
```
1234.560059
1234.560059
```

Output in broken case:
```
1234.560059
0.0
```

Reviewers: #powerpc

Subscribers: jhibbits, luporl, alfredo

Tags: #powerpc

Differential Revision: https://reviews.freebsd.org/D29851
cgit
b0025f9b emaste Sept. 4, 2021, 3:33 p.m.
Fixes:		2f513db72b03 ("Upgrade to OpenSSH 7.9p1.")
MFC after:	3 days
Sponsored by:	The FreeBSD Foundation
cgit
846a6e8f kp Sept. 4, 2021, 2:07 p.m.
795d78a46713 pfctl: Don't use PRIu64 mistakenly changed these to be
printed as hexadecimal numbers.

Reported by:	Florian Smeets
MFC after:	4 days
Sponsored by:	Rubicon Communications, LLC ("Netgate")
cgit
88a3af4d gbe Sept. 4, 2021, 11:02 a.m.
- s/valiables/variables/

Obtained from:	NetBSD
MFC after:	3 days
cgit