Commit Graph

243606 Commits

Author SHA1 Message Date
msaitoh
d78e3b9a5a Add gpioiic into amd64/conf/ALL 2016-07-11 11:25:54 +00:00
skrll
83e52e0af5 Don't use a callout unnecessarily - especially not this early 2016-07-11 10:55:35 +00:00
wiz
6c143561b1 Bump date for function prototype change. 2016-07-11 10:47:52 +00:00
knakahara
53a51f2a77 strncpy should use destination buf length instead of source buf length.
pointed out by nonaka@n.o.
2016-07-11 09:42:20 +00:00
skrll
16e89468ad Fix an aprint_normal 2016-07-11 08:09:15 +00:00
msaitoh
6399f1a6ef KNF. No functional change. 2016-07-11 07:42:13 +00:00
ozaki-r
dca032f9f4 Run timers in workqueue
Timers (such as nd6_timer) typically free/destroy some data in callout
(softint). If we apply psz/psref for such data, we cannot do free/destroy
process in there because synchronization of psz/psref cannot be used in
softint. So run timer callbacks in workqueue works (normal LWP context).

Doing workqueue_enqueue a work twice (i.e., call workqueue_enqueue before
a previous task is scheduled) isn't allowed. For nd6_timer and
rt_timer_timer, this doesn't happen because callout_reset is called only
from workqueue's work. OTOH, ip{,6}flow_slowtimo's callout can be called
before its work starts and completes because the callout is periodically
called regardless of completion of the work. To avoid such a situation,
add a flag for each protocol; the flag is set true when a work is
enqueued and set false after the work finished. workqueue_enqueue is
called only if the flag is false.

Proposed on tech-net and tech-kern.
2016-07-11 07:37:00 +00:00
msaitoh
186ee438ac - Use aprint*() instead of printf() in the attach function.
- KNF.
2016-07-11 07:11:08 +00:00
knakahara
bf83a4a723 pci_intr_type() is required pci_chipset_tag_t argument by other than x86.
pointed out by nonaka@n.o.
2016-07-11 06:14:51 +00:00
matt
981257b6b6 Add makeoptions NEED_MDSETIMAGE="yes" 2016-07-11 03:45:59 +00:00
ozaki-r
8d7855e3f6 Revert part of "Switch the address list of intefaces to pslist(9)" (r1.220)
Reverting the whole change set just messes up many files uselessly
because changes to them (except for if.h) are proper.

- Remove ifa_pslist_entry that breaks kvm(3) users (e.g., netstat -ia)
- Change IFADDR_{READER,WRITER}_* macros to use old IFADDR_* (or just NOP)
  for now

Fix PR kern/51325
2016-07-11 02:14:27 +00:00
matt
f9a5ff338e Add machdep.fpu_present sysctl node 2016-07-11 01:50:17 +00:00
jmcneill
17dde1d631 fix inverted screen blanking logic 2016-07-10 21:40:45 +00:00
riastradh
de8d1fd320 Make kernel core dumps on cgd(4) work. 2016-07-10 17:40:23 +00:00
pgoyette
f7331529dc The piixpm(4) driver also supports AMD's HUDSON chipset's SMBus function.
From OpenBSD, verified.
2016-07-10 04:44:47 +00:00
pgoyette
1551b030ab Don't assign value to rdev unless we're going to use the value.
Therefore, don't declare the variable unless we're going to assign it a
value.
2016-07-10 01:35:27 +00:00
jmcneill
582b02e9f7 The version of PMON on my 8089B (LM8089-1.4.9a) configures OHCI but doesn't
enable memory space access. Enable it before the ohci driver attaches to
allow low- and full-speed devices to work.
2016-07-10 00:14:36 +00:00
dholland
ff45d31ff1 bump up the copyright date since I practically rewrote the file 2016-07-09 23:13:47 +00:00
dholland
bb5b28e611 Document all the signals properly. Please review.
Kick out the list of async-signal-safe functions (it is now in
sigaction(2)) because this is not a programmer-facing document.

XXX: if anyone knows definitively whether SIGEMT is actually used by
XXX: anything, please update the weasel words I put in.
2016-07-09 23:12:27 +00:00
dholland
20b32a3eee Drop in a reference to the list of async-signal-safe functions. 2016-07-09 23:10:05 +00:00
dholland
b00b33da18 Dump the list of async-signal-safe functions here. It belongs in
programmer-facing docs, not user-facing docs, so shouldn't be in
signal(7).
2016-07-09 23:08:32 +00:00
dholland
3944777ad3 Fix three of these strings (ones that are rarely seen) 2016-07-09 21:15:00 +00:00
dholland
323c08e328 note that on libc bump obsolete signal functions should be sent to
libcompat.
2016-07-09 20:15:06 +00:00
kiyohara
113cfd1e52 Remove white-spaces. 2016-07-09 15:04:06 +00:00
maxv
8deb66a194 Simplify pmap_get_physpage. 2016-07-09 09:33:21 +00:00
maxv
6d9a9c0bf4 Use pmap_bootstrap_palloc. 2016-07-09 09:25:44 +00:00
maxv
066b39e9dc When a user pmap is created, it is populated with the higher kernel
slots, which become accessible upon kernel entry (syscall, cpu switch,
or whatever). Put the NOX bit in the user recursive slot, so the whole
tree does not appear as executable in kernel mode.

This is already what is done in the kernel pmap.
2016-07-09 08:05:46 +00:00
maxv
4fb21ac06f KNF this function a little 2016-07-09 07:47:25 +00:00
maxv
6c1bb9a544 When loading a module from VFS and from the bootloader, the kernel packs
up the module segments into one big RWX chunk. Split this chunk into two
different text and data+bss+rodata chunks. The latter is made non-
executable. This also provides some kind of ASLR, since the chunks are
not necessarily contiguous.
2016-07-09 07:25:00 +00:00
maxv
e6085ad1d6 The CPU considers a given va as executable if none of its levels have
the NOX bit. With the top level recursive slot, however, several levels
are recursively omitted, which implies that each entry that is not the
child of NOX-ed parents actually appears somewhere in the virtual space
as executable via this slot, even if it is followed by an underlying
entry that has the NOX bit.

This recursive slot is only used to edit the page tree itself. Make it
non-executable.
2016-07-09 06:58:06 +00:00
skrll
c3cc24a21b KNF 2016-07-09 06:49:03 +00:00
matt
8a31a6e7a0 Pass down ${CPUFLAGS} 2016-07-09 05:52:22 +00:00
macallan
232bfe5631 first step to make genfb work with /chaos/control - don't do config space
writes on /chaos.
tested by jak
2016-07-08 22:21:52 +00:00
maxv
c7e8a1a390 The preloaded modules are now reallocated dynamically by the kernel. This
area does not need to be executable anymore.
2016-07-08 09:15:38 +00:00
maxv
e169fdcc18 Force the kernel to dynamically reallocate the preloaded modules. 2016-07-08 08:55:48 +00:00
ozaki-r
bdcbfcd17f Fix test names 2016-07-08 08:27:07 +00:00
skrll
a867508b57 Remove '\n' from UVMHIST_LOG format - it is not needed. 2016-07-08 06:45:34 +00:00
ozaki-r
94dba1b837 CID 1363345: remove unreachable code and cleanup returns 2016-07-08 06:18:29 +00:00
ozaki-r
6e4af1673a CID 1363344: remove dead code
We may need to reconsider a case when m_get_rcvif_psref returns NULL.
2016-07-08 06:15:33 +00:00
skrll
b61c1800b7 Don't ignore return value of xhci_update_ep0_mps and remove KASSERT at
the end of that function.

From t-hash
2016-07-08 05:38:31 +00:00
skrll
f062101a0a More debug in new_device.
From t-hash.
2016-07-08 05:37:38 +00:00
skrll
96197f702e s/UBS/USB/
From t-hash
2016-07-08 05:36:51 +00:00
ozaki-r
4133a8eca8 Replace macros to get an IP address with proper inline functions
The inline functions are more friendly for applying psz/psref;
they consist of only simple interations.
2016-07-08 04:33:30 +00:00
ozaki-r
75a23513d7 Kill remaining use of the old lists of IP addresses 2016-07-08 03:40:34 +00:00
christos
abddfef7f8 more fixes. 2016-07-08 01:24:53 +00:00
matt
1c4b7019be Add placeholder for compiing earmv7* with -mthumb.
Preserve $d/$a/$t for all ARM binaries regardless of endian.
2016-07-07 20:52:53 +00:00
kiyohara
bbc92c4d2e com@obio's type is COM_TYPE_OMAP. This mean is able to change the baudrate.
tested on VTC100's com1(GPS).
2016-07-07 15:27:42 +00:00
kiyohara
9161291044 Call uart_enable() before comprobe1().
And rename to uart_enable_omap().  Also do soft reset in new uart_enable().
Add uart_enable_am335x().  Its enable to clocks.
2016-07-07 15:20:58 +00:00
maya
44573f1093 Turn a few KASSERTs into KASSERTMSG mentioning the tested values
Some of these values are optimized out, so knowing why the test failed
can be challenging.
2016-07-07 10:53:03 +00:00
ozaki-r
9e4c2bda8a Switch the address list of intefaces to pslist(9)
As usual, we leave the old list to avoid breaking kvm(3) users.
2016-07-07 09:32:01 +00:00