jhb [Tue, 13 Jan 2009 16:44:18 +0000 (16:44 +0000)]
Add support for two new event source types:
- Callwheels traced via KTR_CALLOUT. Each CPU is assigned a callwheel
source. The events on this source are the execution of individual callout
routines. Each routine shows up as a green rectangle while it is executed
and the event details include the function pointer and argument.
- Locks traced via KTR_LOCK. Currently, each lock name is assigned an event
source (since the existing KTR_LOCK traces only include lock names and
not pointers). This does mean that if multiple locks of the same name are
manipulated, the source line for that name may be confusing. However, for
many cases this can be useful. Locks are blue when they are held and
purple when contested. The contention support is a bit weak due to
limitations in the rw_rlock() and mtx_lock_spin() logging messages
currently. I also have not added support for contention on lockmgr,
sx, or rmlocks yet. What is there now can be profitably used to examine
activity on Giant however.
- Expand the width of the event source names column a bit to allow for some
of the longer names of these new source types.
jhb [Tue, 13 Jan 2009 16:33:10 +0000 (16:33 +0000)]
- Add some rudimentary support for sorting the list of event sources
(threads, CPU load counters, etc.). Each source is tagged with a group
and an order similar to the SYSINIT SI_SUB_* and SI_ORDER_*. After
the file is parsed, all the sources are then sorted. Currently, the only
affects of this are that the CPU loads are now sorted by CPU ID (so
CPU 0 is always first). However, this makes it easier to add new types
of event sources in the future and have them all clustered together
instead of intertwined with threads.
- Python lists perform insertions at the tail much faster than insertions
at the head. For a trace that had a lot of events for a single event
source, the constant insertions of new events to the head of the
per-source event list caused a noticable slow down. To compensate,
append new events to the end of the list during parsing and then
reverse the list prior to drawing.
- Somewhere in the tkinter internals the coordinates of a canvas are
stored in a signed 32-bit integer. As a result, if an the box for
an event spans 2^31, it would actually end up having a negative
X offset at one end. The result was a single box that covered the
entire event source. Kris worked around this for some traces by
bumping up the initial ticks/pixel ratio from 1 to 10. However, a
divisor of 10 can still be too small for large tracefiles (e.g.
with 4 million entries). Instead of hardcoding the initial scaling
ratio, calculate it from the time span of the trace file.
- Add support for using the mouse wheel to scroll the graph window
up and down.
mav [Tue, 13 Jan 2009 16:27:04 +0000 (16:27 +0000)]
Change configuration order to enable output only after codec is configured.
Mute all mixer controllable amplifiers initially to let mixer to unmute
only some of them later. This should reduce clicks and noises during boot.
raj [Tue, 13 Jan 2009 15:41:58 +0000 (15:41 +0000)]
Rework BookE pmap towards multi-core support.
o Eliminate tlb0[] (a s/w copy of TLB0)
- The table contents cannot be maintained reliably in multiple MMU
environments, where asynchronous events (invalidations from other cores)
can change our local TLB0 contents underneath.
- Simplify and optimize TLB flushing: system wide invalidations are
performed using tlbivax instruction (propagates to other cores), for
local MMU invalidations a new optimized routine (assembly) is introduced.
o Improve and simplify TID allocation and management.
- Let each core keep track of its TID allocations.
- Simplify TID recycling, eliminate dead code.
- Drop the now unused powerpc/booke/support.S file.
obrien [Tue, 13 Jan 2009 06:52:51 +0000 (06:52 +0000)]
Set .POSIX mode, along with disable parallel mode. Crunchgen parses the
output from make(1) and its picky what it should look like.
Also use make's conditional assignment operator rather than test and set.
sam [Tue, 13 Jan 2009 05:50:22 +0000 (05:50 +0000)]
On some platforms touching the bb registers when the phy is powered
down will cause a fault. Check the phy power state before possibly
reading from the bb, this can happen as ar5212Reset intentionally
calls ar5212GetRfgain before bringing the bb out of reset (but we
do it here and not in the caller to guard against other possible uses).
obrien [Mon, 12 Jan 2009 21:49:42 +0000 (21:49 +0000)]
If running with "-K" really do not information about symbols from the a.out
argument. Before this fix, after searching the currently-running kernel,
we would still search the a.out argument - completely override the in-kernel
list, essentially defeating the K flag's purpose.
PR: 47387
Submitted by: Ryan Beasley <ryanb@goddamnbastard.org>
obrien [Mon, 12 Jan 2009 17:18:58 +0000 (17:18 +0000)]
Fix issue where ata_atapicmd() can never really return EBUSY which is
expected in acd_fixate().
This should fix various problems folks are having with 'burncd' reporting
"burncd: ioctl(CDRIOCFIXATE): Input/output error" during the fixate phase
when "fixate" is issued together with the "data" command.
PR: 95979
Submitted by: Jaakko Heinonen <jh@saunalahti.fi>
qingli [Mon, 12 Jan 2009 11:24:32 +0000 (11:24 +0000)]
Revive the RTF_LLINFO flag in route.h. The kernel code is guarded
by the new kernel option COMPAT_ROUTE_FLAGS for binary backward
compatibility. The RTF_LLDATA flag maps to the same value as RTF_LLINFO.
RTF_LLDATA is used by the arp and ndp utilities. The RTF_LLDATA flag is
always returned to the userland regardless whether the COMPAT_ROUTE_FLAGS
is defined.
nwhitehorn [Mon, 12 Jan 2009 01:36:01 +0000 (01:36 +0000)]
Some early Macintosh GPIO controllers don't provide reg properties for
interrupt-only GPIOs. Honor this, and allow interrupt attachment, but not
read/write access for such devices.
rwatson [Sat, 10 Jan 2009 23:50:23 +0000 (23:50 +0000)]
Do invoke mac_ifnet_check_transmit() and mac_ifnet_create_mbuf()
in the loopback and synthetic loopback code so that packets are
access control checked and relabeled. Previously, the MAC
Framework enforced that packets sent over the loopback weren't
relabeled, but this will allow policies to make explicit choices
about how and whether to relabel packets on the loopback. Also,
for SIMPLEX devices, this produces more consistent behavior for
looped back packets to the local MAC address by labeling those
packets as coming from the interface.
trasz [Sat, 10 Jan 2009 17:22:49 +0000 (17:22 +0000)]
Don't call destroy_dev(9) with a mutex held. While here, shuffle
things around so the periph destructors look alike. Based on a patch
by Jaakko Heinonen.
Submitted by: Jaakko Heinonen
Reviewed by: scottl
Approved by: rwatson (mentor)
Sponsored by: FreeBSD Foundation
rwatson [Sat, 10 Jan 2009 10:58:41 +0000 (10:58 +0000)]
Rather than having MAC policies explicitly declare what object types
they label, derive that information implicitly from the set of label
initializers in their policy operations set. This avoids a possible
class of programmer errors, while retaining the structure that
allows us to avoid allocating labels for objects that don't need
them. As before, we regenerate a global mask of labeled objects
each time a policy is loaded or unloaded, stored in mac_labeled.
Discussed with: csjp
Suggested by: Jacques Vidrine <nectar at apple.com>
Obtained from: TrustedBSD Project
Sponsored by: Apple, Inc.
Restrict arp, ndp and theoretically the FIB listing (if not
read with libkvm) to the addresses of a prison, when inside a
jail. [1]
As the patch from the PR was pre-'new-arp', add checks to the
llt_dump handlers as well.
While touching RTM_GET in route_output(), consistently use
curthread credentials rather than the creds from the socket
there. [2]
adrian [Fri, 9 Jan 2009 16:02:19 +0000 (16:02 +0000)]
Implement a new IP option (not compiled/enabled by default) to allow
applications to specify a non-local IP address when bind()'ing a socket
to a local endpoint.
This allows applications to spoof the client IP address of connections
if (obviously!) they somehow are able to receive the traffic normally
destined to said clients.
This patch doesn't include any changes to ipfw or the bridging code to
redirect the client traffic through the PCB checks so TCP gets a shot
at it. The normal behaviour is that packets with a non-local destination
IP address are not handled locally. This can be dealth with some IPFW hackery;
modifications to IPFW to make this less hacky will occur in subsequent
commmits.
Thanks to Julian Elischer and others at Ironport. This work was approved
and donated before Cisco acquired them.
Obtained from: Julian Elischer and others
MFC after: 2 weeks
Make SIOCGIFADDR and related, as well as SIOCGIFADDR_IN6 and related
jail-aware. Up to now we returned the first address of the interface
for SIOCGIFADDR w/o an ifr_addr in the query. This caused problems for
programs querying for an address but running inside a jail, as the
address returned usually did not belong to the jail.
Like for v6, if there was an ifr_addr given on v4, you could probe
for more addresses on the interfaces that you were not allowed to see
from inside a jail. Return an error (EADDRNOTAVAIL) in that case
now unless the address is on the given interface and valid for the
jail.