Ed Maste [Fri, 12 Feb 2016 20:54:02 +0000 (20:54 +0000)]
Update ELF Tool Chain to upstream rev 3400
Some notable improvements include:
readelf:
- Add AArch64 relocation definitions.
- Report value of unknown relocation types.
elfcopy:
- Consider symbols with STB_GNU_UNIQUE binding as global symbols.
- Fixed support for VMA adjustment for loadable sections found
in relocatable objects.
- Handle nameless global symbols.
- Improve wildcard matching for !-prefixed symbols.
- Add PE/COFF support.
Mark Johnston [Fri, 12 Feb 2016 20:43:53 +0000 (20:43 +0000)]
Clear the cookie pointer on error in tmpfs_readdir().
It is otherwise left dangling, and callers that request cookies always free
the cookie buffer, even when VOP_READDIR(9) returns an error. This results
in a double free if tmpfs_readdir() returns an error to the NFS server or
the Linux getdents(2) emulation code.
Andrew Turner [Fri, 12 Feb 2016 12:38:04 +0000 (12:38 +0000)]
Only update curthread and curpcb after we have finished using the old
values.
If switching from a thread that used floating-point registers to a thread
that is still running, but holding the blocked_lock lock we would switch
the curthread to the new (running) thread, then call critical_enter. This
will non-atomically increment td_critnest, and later call critical_exit to
non-atomically decrement this value.
This can happen at the same time as the new thread is still running on the
old core, also calling these functions. In this case there will be a race
between these non-atomic operations. This can be an issue as we could loose
one of these operations leading to the value to not return to zero.
If, later on, we then hit a data abort we check if the td_critnest is zero.
If this check fails we will panic the kernel.
This has been observed when running pcmstat on a Cavium ThunderX. The pcm
thread will use the blocked_lock lock and there is a high chance userspace
will use the floating-point registers. When, later on, pmcstat triggers a
data abort we will hit this panic.
The fix is to update these values after storing the floating-point state.
This means we use the correct curthread while storing the state so it will
not be an issue that the changes to td_critnest are non-atomic.
POSIX states that #include <signal.h> shall make both mcontext_t and
ucontext_t available. Our code even has XXX comment about this.
Add a bit of compliance by moving struct __ucontext definition into
sys/_ucontext.h and including it into signal.h and sys/ucontext.h.
Several machine/ucontext.h headers were changed to use namespace-safe
types (like uint64_t->__uint64_t) to not depend on sys/types.h.
struct __stack_t from sys/signal.h is made always visible in private
namespace to satisfy sys/_ucontext.h requirements.
Apparently mips _types.h pollutes global namespace with f_register_t
type definition. This commit does not try to fix the issue.
PR: 207079
Reported and tested by: Ting-Wei Lan <lantw44@gmail.com>
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
If full width writes to the performance monitoring counters are
supported, use full-width aliases MSRs for writes. This fixes the
"[pmc,X] negative increment" assertion on the context switch when
clipped counter value is sign-extended.
Add definitions for the MSR IA32_PERF_CAPABILITIES needed to detect
the feature.
Devin Teske [Fri, 12 Feb 2016 02:53:44 +0000 (02:53 +0000)]
Comments and fix small bug
Reduce differences between jib/jng and fix a bug that would prevent
additional interfaces from being created if the first of many already
existed (counter wasn't incremented before calling only continue).
Devin Teske [Fri, 12 Feb 2016 01:41:40 +0000 (01:41 +0000)]
Add syntax to disable MAC allocation
Adding `!' before an interface name will disable MAC allocation, falling
back to driver mechanics. Alternatively adding `=' before an interface name
causes the MAC address to be cloned (for ng_bridge(4) back-end only). While
here, disable the auto-detection of wlan* since this knocks the host off;
requiring the host that defines the jail to explicitly enable this feature
by preceding the interface with `='.
Devin Teske [Thu, 11 Feb 2016 22:10:54 +0000 (22:10 +0000)]
Add support for bridging iwn(4) based wlan(4)
Documented in iwn(4), "Only one virtual interface may be configured at any
time." However, netgraph with a cloned MAC address is able to communicate
over an ng_eiface attached to an ng_bridge linked to the wlan(4) interface.
While here, introduce syntax to specify the MAC address is to be cloned if
the named interface begins with equals [=].
Gleb Smirnoff [Thu, 11 Feb 2016 21:32:23 +0000 (21:32 +0000)]
o Gather all mbuf(9) allocation functions into kern_mbuf.c, and all
mbuf(9) manipulation functions into uipc_mbuf.c. This looks like
the initial intent, but had diffused in the last decade.
o Gather all declarations in mbuf.h in one place and sort them.
o Uninline m_clget() and m_cljget().
There are no functional changes in this patch.
The patch comes from a larger version, where all mbuf(9) allocation was
uninlined, which allowed to make mbuf(9) UMA zones private to kern_mbuf.c.
The performance impact of the total uninlining is still unclear, so we
are holding on now with larger version.
Fix double-free error: r289419 moved all error handling in http_connect()
to the end of the function, but did not remove a fetch_close() call which
was made redundant by the one in the shared error-handling code.
PR: 206774
Submitted by: Christian Heckendorf <heckendorfc@gmail.com>
MFC after: 3 days
Jim Harris [Thu, 11 Feb 2016 17:32:41 +0000 (17:32 +0000)]
nvme: avoid duplicate SET_NUM_QUEUES commands
nvme(4) issues a SET_NUM_QUEUES command during device
initialization to ensure enough I/O queues exists for each
of the MSI-X vectors we have allocated. The SET_NUM_QUEUES
command is then issued again during nvme_ctrlr_start(), to
ensure that is properly set after any controller reset.
At least one NVMe drive exists which fails this second
SET_NUM_QUEUES command during device initialization. So
change nvme_ctrlr_start() to only issue its SET_NUM_QUEUES
command when it is coming out of a reset - avoiding the
duplicate SET_NUM_QUEUES during device initialization.
Reported by: gallatin
MFC after: 3 days
Sponsored by: Intel
Justin Hibbits [Thu, 11 Feb 2016 13:15:37 +0000 (13:15 +0000)]
Migrate the PTE format for book-e to standardize on the 'indirect PTE' format
Summary:
The revised Book-E spec, adding the specification for the MMUv2 and e6500,
includes a hardware PTE layout for indirect page tables. In order to support
this in the future, migrate the PTE format to match the MMUv2 hardware PTE
format.
Test Plan: Boot tested on a P5020 board. Booted to multiuser mode.
Zbigniew Bodek [Thu, 11 Feb 2016 12:04:58 +0000 (12:04 +0000)]
Support interrupts binding in GICv3 and ITS
- Add MOVI command and routine for the LPI migration
- Allow to search for the ITS device descriptor using
not only devID but also LPI number.
- Bind SPIs in the Distributor
- Don't bind its_dev to collection. Keep track of the collection
IDs for each LPI.
Zbigniew Bodek [Thu, 11 Feb 2016 11:59:32 +0000 (11:59 +0000)]
Call pmc_hook() correctly in the ARM64 interrupt handler
pmc_hook() was called only in case of the stray interrupt but should
rather be called on each interrupt. Move in to the arm_cpu_intr()
handler, out of the critical section too.
Zbigniew Bodek [Thu, 11 Feb 2016 11:57:13 +0000 (11:57 +0000)]
Fix bugs in interrupts allocation on ARM64
Separate interrupt descriptors lookup from allocation. It was possible
to perform config on non-existing interrupt simply by allocating spurious
descriptor.
Must lock the interrupt descriptors table lookup to avoid mismatches.
This ought to prevent trouble while setting up new interrupt
and dispatching existing one.
Use spin mutex rather than sleep mutex. This is mainly due to lock in
arm_dispatch_intr.
This should be eventually changed to a lock-less solution without
walking through a linked list on each interrupt.
Roger Pau Monné [Thu, 11 Feb 2016 11:50:31 +0000 (11:50 +0000)]
xen-netfront: remove pointless assignment in xn_ioctl
The variable error is assigned to 0 before entering the switch.
Assigning error to 0 before break pointless rewrites the real error
value that should be returned.
Coverity ID: 1304974
Submitted by: Wei Liu <wei.liu2@citrix.com>
Reviewed by: royger
Sponsored by: Citrix Systems R&D
Differential revision: https://reviews.freebsd.org/D5250
Use a pair of ifs when comparing the 32-bit flowid integers so that
the sign bit doesn't cause an overflow. The overflow manifests itself
as a sorting index wrap around in the middle of the sorted array,
which is not a problem for the LRO code, but might be a problem for
the logic inside qsort().
Adrian Chadd [Thu, 11 Feb 2016 06:09:27 +0000 (06:09 +0000)]
Begin the MIPS_INTRNG support.
This is a prelude to intr-ng support for MIPS boards that need it -
notably the CI20 port from kan@ that's upcoming, but also work that
Stanislav is doing for the Mediatek platforms.
This is the initial platform dependent bits in include/intr.h, some
#defines for the nexus code for the intrng initialisation/runtime
bits, some changed naming (which I'll fix later to be the same, much
like what I did for ARM intr-ng) in exception.S, and the first cut
at a PIC.
Stanislav and I refactored out the common code for intrng support,
so the mips intrng definitions are quite small (sys/mips/include/intr.h.)
This is all work done by kan@, which stanislav has been cherry picking
into common code for his mediatek chipset work.
Tested:
* Carambola2 - no regressions (not intr-ng though!)
Submitted by: Stanislav Galabov <sgalabov@gmail.com>
Reviewed by: kan (original author)
Differential Revision: https://reviews.freebsd.org/D5182
Pedro F. Giffuni [Thu, 11 Feb 2016 00:34:11 +0000 (00:34 +0000)]
Ext4: fix handling of files with sparse blocks before extent's index.
This is ongoing work from Damjan Jovanovic to improve ext4 read support
with sparse files:
Keep track of the first and last block in each extent as it descends down
the extent tree, thus being able to work out that some blocks are sparse
earlier. This solves an issue on r293680.
In ext4_bmapext() start supporting the runb parameter, which appears to be
the number of adjacent blocks prior to the block being converted in the
same way that runp is the number of blocks after, speding up random access
to mmaped files.
Conrad Meyer [Wed, 10 Feb 2016 20:28:28 +0000 (20:28 +0000)]
ntb_hw(4): Allow any x86 PAT caching flags for MW defaults
Replace the hw.ntb.enable_writecombine tunable with
hw.ntb.default_mw_pat. It can be set with several specific numerical
values to select a caching type. Any bogus value is treated as
Uncacheable (UC).
The ntb_mw_set_wc() KPI has removed the restriction that the selected
mode must be one of UC, WC, or WB.
Enji Cooper [Wed, 10 Feb 2016 17:57:24 +0000 (17:57 +0000)]
Similar to r295116, add an additional 1 second sleep after calling ggatel
before calling dd to defeat a race when writing out to the geom_gate(4)
device
Mark Felder [Wed, 10 Feb 2016 16:13:59 +0000 (16:13 +0000)]
Add new rc.conf parameter "jail_reverse_stop"
When a user defines "jail_list" in rc.conf the jails are started in the
order defined. Currently the jails are not are stopped in reverse order
which may break dependencies between jails/services and prevent a clean
shutdown. The new parameter "jail_reverse_stop" will shutdown jails in
"jail_list" in reverse order when set to "YES".
Please note that this does not affect manual invocation of the jail rc
script. If a user runs the command
# service jail stop jail1 jail2 jail3
the jails will be stopped in exactly the order specified regardless of
jail_reverse_stop being defined in rc.conf.
Andrew Rybchenko [Wed, 10 Feb 2016 12:14:56 +0000 (12:14 +0000)]
sfxge: implement SIOCGI2C to read information from phy modules
The IOCTL is used by 'ifconfig -v' to show SFP+/QSFP+ information
including inventory information and dianostics (temperature, light
levels, voltage etc).
Reviewed by: gnn,melifaro
Sponsored by: Solarflare Communications, Inc.
MFC after: 2 days
Differential Revision: https://reviews.freebsd.org/D5240
Stefan Eßer [Wed, 10 Feb 2016 09:40:45 +0000 (09:40 +0000)]
Remove O_SYNC from the options passed to dbmopen().
The services db is created as a temporary file that is moved over the
existing file after completion. Thus there is no need to immediately
flush all created db records to the temporary file.
This speeds up creation of the services db by a factor of 500 on my
ZFS based /var/db filesytem (from 110 seconds to 0.235 seconds).
Andrew Turner [Wed, 10 Feb 2016 09:19:29 +0000 (09:19 +0000)]
Update of the Allwinner drivers to:
* Use the Linux compat string
* Use EARLY_DRIVER_MODULE to attach at the right time
* Add a generic A10 kernel config file
* A20 now use generic_timer
* Add two new dts files for Olimex boards
* Update our custom DTS file for A10 and A20 to use the same compatible
property names as the vendor ones.
Kurt Lidl [Tue, 9 Feb 2016 21:00:38 +0000 (21:00 +0000)]
Zero pf rule counters so daily reports make sense
Zero pf rule counters so that each daily report lists an absolute
number of rejected packets, not the total since the last time the
machine rebooted (or the counters were manually cleared).