0d6e9da228
Make {u,}intptr_t "long" on both sparc and sparc64 (as the NetBSD headers
...
always did)
2014-08-15 09:25:25 +00:00
b9e7c24bbc
Fix the SCN*FAST* defines
2014-08-15 08:59:58 +00:00
adeb9b3e44
Fix editor mishap
2014-08-15 07:53:37 +00:00
3cc0e3b686
All FAST datatypes are int (32 bit) on both sparc and sparc64, only FAST64
...
variants are long (sparc64) or long long (sparc).
2014-08-15 07:45:31 +00:00
e1e8d3be37
avoid calling into time code when cold, and avoid calling nanotime()
...
if we're not going to use the result anyway.
XXX: not necessarily the best fix, but better than crashing *early*
boot due to too-early nanotime() calls.
XXX: pullup.
2014-08-15 07:39:25 +00:00
01186a294e
Do the wildcard symbol dance for the debug case as well.
2014-08-14 20:59:54 +00:00
6faaf3e4db
Use wildcards for stripping/preserving the mapping symbols on ARM and
...
AArch64. LLVM creates unique symbols in each file of the form $a.n etc.
2014-08-14 18:39:38 +00:00
fb363505f8
LLVM sshramdisk needs 8.9MB, so bump again.
2014-08-14 18:38:29 +00:00
709fefda49
http://m00nbsd.net/ae123a9bae03f7dde5c6d654412daf5a.html#Report-2
...
#06-0x01: Empty compiler block
ok christos@
2014-08-14 17:29:30 +00:00
31fca0faec
Remove tautologies.
2014-08-14 16:55:02 +00:00
76be0e7906
Zero ttm_agp objects on creation.
2014-08-14 16:50:22 +00:00
1a58402320
Use ttm_dma_tt_init in ttm_agp_tt_create so we can use ttm_bus_dma.
2014-08-14 16:47:52 +00:00
e451d25590
Bump ramdisk size (since we added loads of firmware to it)
2014-08-14 16:32:11 +00:00
2e29e09879
Restore placement of percpu_init in rump_init.
...
Probably doesn't matter, but let's avoid needless churn around the
real bug fix.
2014-08-14 16:28:49 +00:00
b018e9788e
Lock cprng->cs_lock around rndsink_request to avoid race with callback.
2014-08-14 16:28:30 +00:00
7e616d7734
Defer cprng_fast_init until CPUs are detected.
2014-08-14 16:27:55 +00:00
75f2a57740
CD images need libarchive if MKBSDTAR is yes.
2014-08-14 16:27:17 +00:00
0d53576a4b
Add dtrace modules to amd64-xen module set list. Hi jnemeth!
2014-08-14 16:26:21 +00:00
1ade0b5b66
Add tar.debug and cpio.debug to lists for MKBSDTAR=yes MKDEBUG=yes.
2014-08-14 16:25:20 +00:00
bcba9fe629
Allow bits to be used for the usb hub on Odroid-XU
2014-08-14 15:57:51 +00:00
7138b5e964
Add some examples of machine-specific foo/boot.8 man pages.
2014-08-14 15:50:42 +00:00
2b9dfdd641
Add exyogpio device attachment for exynos5
2014-08-14 15:46:57 +00:00
3453b7f9d9
More detail about treating a local file as a man page,
...
and about how to request machine-specific man pages.
re
2014-08-14 15:44:47 +00:00
727a52919d
Add gpio not connected definitions for Exynos5
...
While here, increase delay around lan9730 powercycle timing to improve its
reset.
2014-08-14 15:44:42 +00:00
7d1acb2eff
Add GPIO definitions for Exynos5 series (taken from 5250)
2014-08-14 15:42:31 +00:00
83de1b41be
Rename C2C gpio to match the naming scheme
...
Add I2S gpio offset
2014-08-14 15:41:07 +00:00
97f5942bd4
Adjust AUDIOCORE mapping for exynos4 to map in entire piece.
2014-08-14 15:39:31 +00:00
ee28f0e59d
For an argument to be interpreted as a local file name, bypassing the
...
search rules in man.conf or MANPATH, it must begin with "/", "./", or
"../". Simply testing whether it contains "/" is wrong, because it
breaks usage like "man 8 vax/boot".
This reverts revision 1.57 dated 2013-10-06,
"Be more permissive in interpreting man pages as filenames".
2014-08-14 15:31:12 +00:00
13fd61cc78
Fix INTPTR_TYPE/UINTPTR_TYPE for n32.
2014-08-14 15:16:20 +00:00
c8bf6ef46e
Fix typo in OBJCOPYLIBFLAGS_EXTRA for aarch64eb.
2014-08-14 15:14:45 +00:00
0e582fedb6
Fix SCN?FAST{8,16} formats to match reality (int) in gcc4.8.3
2014-08-14 14:21:03 +00:00
b0e4602c25
Overflow if *data_len == OSIZE and args->version >= PTYFS_ARGSVERSION.
...
Sent on tech-kern@, ok christos@
2014-08-14 14:06:53 +00:00
6148f733b8
Set __NO_FIQ for the Exynos processors are shipped with SecureCore enabled
...
gics and FIQs are for the secure part only.
2014-08-14 12:48:52 +00:00
e1cf4281db
Force INTPTR_TYPE/UINTPTR_TYPE to long int/unsigned long int.
...
(same types as PTRDIFF_TYPE/SIZE_TYPE)
2014-08-14 11:32:18 +00:00
9842fff13a
Revert previous changes, gcc configuration will be adapted instead.
2014-08-14 11:28:13 +00:00
b209a3f1b9
intptr_t and uintptr_t are not long any more.
2014-08-14 09:17:32 +00:00
66ee8b8898
Previously we used unsigned long int as __uintptr_t, but gcc prefers
...
__UINTPTR_TYPE__ as unsigned int, and now we use that for __uintptr_t
as well, so adapt MUTEX_CAS() accordingly.
2014-08-14 09:02:18 +00:00
3722b36159
Set ifflags callback so that the device can enter promiscuous mode.
2014-08-14 05:42:16 +00:00
b88adb6f3e
Add missing delay slot in DEBUG kernel.
2014-08-14 04:14:33 +00:00
82a12988ef
Include <sys/common_int_types.h> if __UINTPTR_TYPE__ is defined.
2014-08-13 22:56:56 +00:00
0b2cac531b
Include <sys/common_int_const.h> if __INTMAX_C_SUFFIX__ is defined.
2014-08-13 22:51:58 +00:00
feeca9ec0c
mention NV_TEST_FOR_KERNEL_DRIVER and explain why it isn't enabled for now
2014-08-13 22:48:15 +00:00
7c5094c243
Include <sys/common_init_limits.h> if __SIG_ATOMIC_MAX__ is defined.
2014-08-13 22:31:06 +00:00
2534797f07
include <sys/common_int_mwgwtypes.h> if __UINT_FAST64_TYPE__ is defined.
2014-08-13 22:25:39 +00:00
ca4be7acf3
Don't save r4 since we don't use it.
2014-08-13 22:16:24 +00:00
c97e21f563
Add __NO_FIQ to prevent testing of the PSR FIQ bit
...
Use VALID_R15_PSR(pc, spsr) instead of testing explicitly
2014-08-13 21:41:32 +00:00
f106fcdb01
When creating a mcontext, make sure the PSR we are saving is actually valid.
2014-08-13 21:10:31 +00:00
22405d4653
Fix shifts & masks in Linux pci_read_config_{word,byte}.
2014-08-13 20:56:21 +00:00
d02c242c7b
Bump ramdisk size to 3.2MB and include dhcpcd.
2014-08-13 20:52:05 +00:00
a54e6af460
Use compiler-rt and libc++ with LLVM on PowerPC64.
2014-08-13 20:51:26 +00:00
martin