markj [Tue, 2 May 2017 23:29:42 +0000 (23:29 +0000)]
Synchronize unclean mirrors before adding them to a running gmirror.
During gmirror startup, if component mirrors are found to be dirty as is
typical after a system crash, the mirrors are synchronized to the mirror
with highest priority. However if a gmirror starts without all of its
mirrors present, for example because of some transient delays during
tasting, the remaining mirrors must be synchronized before they may become
active.
jilles [Tue, 2 May 2017 21:56:20 +0000 (21:56 +0000)]
csplit: Fix check of fputs() return value, making csplit work again.
As of r295638, fputs() returns the number of bytes written (if not more than
INT_MAX). This broke csplit completely, since csplit assumed only success
only for the return value 0.
brooks [Tue, 2 May 2017 21:20:27 +0000 (21:20 +0000)]
Correct an out-of-bounds read in regcomp when the RE is bad.
When passed the invalid regular expression "a**", the error is
eventually detected and seterr() is called. It sets p->error
appropriatly and p->next and p->end to nuls which is a never used char
nuls[10] which is zeros due to .bss initialization. Unfortunatly,
p_ere_exp() and p_simp_re() both have fall through cases where they set
the error, decrement p->next and access it which means a read from what
ever .bss variable comes before nuls.
Found with regex_test:repet_multi and CHERI bounds checking.
emaste [Tue, 2 May 2017 20:44:06 +0000 (20:44 +0000)]
bsdgrep: fix escape map building for multibyte strings
In BSD grep, fix escape map building in the regex parser. It was
previously using memory not explicitly initialized, and the MBS escape
map was being built based on a version of the pattern with escapes
already parsed out.
This is Kyle's change, but I restored the broken style that already
exists in this file.
emaste [Tue, 2 May 2017 20:39:33 +0000 (20:39 +0000)]
bsdgrep: fix -w flag matching with an empty pattern
-w flag matching with an empty pattern was generally 'broken', allowing
matches to occur on any line whether or not it actually matches -w
criteria.
This fix required a good amount of refactoring to address. procline()
is altered to *only* process the line and return whether it was a match
or not, necessary to be able to short-circuit the whole function in case
of this matchall flag. -m flag handling is moved out as well because it
suffers from the same fate as context handling if we bypass any actual
pattern matching.
The matching context (matches, mostly) didn't previously exist outside
of procline(), so we go ahead and create context object for file
processing bits to pass around. grep_printline() was created due to
this, for the scenarios where the matches don't actually matter and we
just want to print a line or two, a la flushing the context queue and
no -o or --color specified.
Damage from this broken behavior would have been mitigated by the fact
that it is unlikely users would invoke grep -w with an empty pattern.
This was identified while checking PR 105221 for problems it this may
cause in BSD grep, but PR 105221 is *not* a report of this behavior.
Submitted by: Kyle Evans <kevans91 at ksu.edu>
Differential Revision: https://reviews.freebsd.org/D10433
cxgbe(4): Support routines for Tx traffic scheduling.
- Create a new file, t4_sched.c, and move all of the code related to
traffic management from t4_main.c and t4_sge.c to this file.
- Track both Channel Rate Limiter (ch_rl) and Class Rate Limiter (cl_rl)
parameters in the PF driver.
- Initialize all the cl_rl limiters with somewhat arbitrary default
rates and provide routines to update them on the fly.
- Provide routines to reserve and release traffic classes.
mav [Tue, 2 May 2017 19:09:11 +0000 (19:09 +0000)]
Introduce sleepable locks into if_lagg.
Before this change if_lagg was using nonsleepable rmlocks to protect its
internal state. This patch introduces another sx lock to protect code
paths that require sleeping, while still uses old rmlock to protect hot
nonsleepable data paths.
This change allows to remove taskqueue decoupling used before to change
interface addresses without holding the lock. Instead it uses sx lock to
protect direct if_ioctl() calls.
As another bonus, the new code synchronizes enabled capabilities of member
interfaces, and allows to control them with ifconfig laggX, that was
impossible before. This part should fix interoperation with if_bridge,
that may need to disable some capabilities, such as TXCSUM or LRO, to allow
bridging with noncapable interfaces.
emaste [Tue, 2 May 2017 18:37:11 +0000 (18:37 +0000)]
bsdgrep: revise test case which will soon become a failure
Work in progress (D10315) is going to make egrep_empty_invalid an
actually invalid regex, to be consistent with the equivalent BRE "{"
behavior, when using regex(3).
Any non-0 exit value is acceptable, depending on how the installed grep
interprets the expression. GNU grep interprets it as non-matching, and
in the future BSD grep will interpret it is an error.
emaste [Tue, 2 May 2017 02:32:10 +0000 (02:32 +0000)]
bsdgrep: fix -w -v matching improperly with certain patterns
-w and -v flag matching was mostly functional but had some minor
problems:
1. -w flag processing only allowed one iteration through pattern
matching on a line. This was problematic if one pattern could match
more than once, or if there were multiple patterns and the earliest/
longest match was not the most ideal, and
2. Previous work "fixed" things to not further process a line if the
first iteration through patterns produced no matches. This is clearly
wrong if we're dealing with the more restrictive -w matching.
#2 breakage could have also occurred before recent broad rewrites, but
it would be more arbitrary based on input patterns as to whether or not
it actually affected things.
Fix both of these by forcing a retry of the patterns after advancing
just past the start of the first match if we're doing more restrictive
-w matching and we didn't get any hits to start with. Also move -v flag
processing outside of the loop so that we have a greater change to match
in the more restrictive cases. This wasn't strictly wrong, but it could
be a little more error prone.
While here, introduce some regressions tests for this behavior and fix
some excessive wrapping nearby that hindered readability. GNU grep
passes these new tests.
Properly initialize ipfw_range_tlv variable to fix possible EINVAL
in case when ipfw delete/zero/resetlog command issued for several rules
in the loop. Also reorder some variables by size.
brooks [Mon, 1 May 2017 20:04:07 +0000 (20:04 +0000)]
Support clnt_raw's use of FD_SETSIZE as a fake file descriptor.
Accomplish this by allocating space for it in __svc_xports and allowing
it to be registered. The failure to allocate space was causing an
out-of-bounds read in svc_getreq_common(). The failure to register
caused PR 211804.
bdrewery [Mon, 1 May 2017 19:34:15 +0000 (19:34 +0000)]
Redo r288270: Hookup mkcsmapper_static and mkesdb_static for 'make clean'
These are only built as part of the top-level 'build-tools' call for
'make buildworld'. They still need to be cleaned during the 'make clean'
treewalks though.
ken [Mon, 1 May 2017 18:53:47 +0000 (18:53 +0000)]
Fix camcontrol timestamp setting and update the man page.
camcontrol timestamp -s would somtimes fail due to stack garbage. Zero out
the timestamp parameters to fix it.
Fix another nearby bug, and update the man page.
sbin/camcontrol/timestamp.c:
In set_timestamp(), bzero ts_p prior to creating the timestamp.
Previously stack garbage could cause some tape drives to reject the
timestamp.
In set_timestamp(), check for failures from strptime().
sbin/camcontrol/camcontrol.8:
Add the time argument to the -T option to camcontrol timestamp -s
in the long description.
Change the time/date format used in the camcontrol timestamp
example to RFC 2822 format. This fixes a time zone issue with the
original example by specifying the time zone as -0600. Otherwise,
the time zone seems to default to standard time in the current
locale, which makes the time, when reported back from the drive,
1 hour off from the intended setting. This also fixes a duplicate
day of the week ("Wednesday Wed") in the previous example.
Submitted by: Sam Klopsch
MFC after: 3 days
Sponsored by: Spectra Logic
tsoome [Mon, 1 May 2017 16:56:34 +0000 (16:56 +0000)]
loader.efi: ResetSystem does not use data with EFI_SUCCESS
The current reboot command in efi/loader/main.c is passing extra data with
ResetSystem, however, UEFI spec 2.6, page 265 does state:
"ResetData is only valid if ResetStatus is something other than EFI_SUCCESS
unless the ResetType is EfiResetPlatformSpecific where a minimum amount of
ResetData is always required."
Therefore we should use DataSize 0 and ResetData NULL - those are two last
arguments for the call.
Restructure normal (non-error) control flow in sem_close().
Do not retest for the found semaphore after the loop to look it up.
Instead, handle both cases of last and non-last close simultaneously,
which allows to consolidate the list unlock and successful return.
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
revert r313473 (Disable LLD_IS_LD option combinations that fail)
r316647 corrected the build of tblgen and libllvm as dependencies for
LLD so undo the temporary seat-belt.
We still want to extend the build infrastructure to automatically detect
the case where the host LLD can be used instead of building a bootstrap
LLD, and likely extend libllvmminimal to meet LLD's needs for cases
where the build includes LLD but not Clang.
When the switch is set to operate in the Multi Chip Addressing Mode we
cannot access the GLOBAL2 register directly.
Despite the comment in code (which was misleading), the indirect access is
only used to read the switch CONFIG data from the scrap register and not
for the GLOBAL2 access.
Use the dsa data to define when the switch is in the Multi Chip Addressing
Mode (a even address different than zero).
printf: Output formatted data directly, instead of via asprintf.
Long ago, sh used to have its own optimized and restricted string formatting
implementation, which the printf builtin had to bypass via asprintf() to a
temporary buffer. Since sh has used libc's string formatting implementation
for a long time, remove the workaround.
Add a check to keep printf %c '' working the same way (output nothing);
POSIX allows both outputting nothing and outputting a NUL byte.
Also, this change avoids silently discarding format directives for whose
output asprintf() cannot allocate memory.
This patch allows the MTU stored in the hostcache to be used as an
initial value for SCTP paths. When an ICMP PTB message is received,
store the MTU in the hostcache.
1. Wrap at <80 columns for readability when editing. Rewrap some lines
prematurely wrapped to better fit in <80 columns and not waste
vertical space.
2. Fix SEE ALSO sorting (sort by section first, then manpage name).
3. Tweak the compound device description slightly by adding soft stops
via commas.
MFC after: 1 week
Reported by: igor [3], manlint [2]
Sponsored by Dell EMC Isilon
marius [Sat, 29 Apr 2017 00:53:17 +0000 (00:53 +0000)]
Fix a bug introduced as part of r287726; use the right device_t for
determining the softc of the bridge in psycho_route_interrupt(). [1]
While at it, update the corresponding comment that the code in
question is also necessary for U30s in addition to E450s (a fact
that has been known for ages).
Modify the NFSv4.1/pNFS client to ask for a maximum length of layout.
The code specified the length of a layout as INT64_MAX instead of
UINT64_MAX. This could result in getting a layout for less than the
full file for extremely large files. Although having little practical
effect, this patch corrects this in the code.
Detected during recent testing of the pNFS server.
cem [Fri, 28 Apr 2017 18:25:10 +0000 (18:25 +0000)]
x86 MCA: Fix a deadlock in MCA exception processing
In exceptional circumstances, an MCA exception will trigger when the
freelist is exhausted. In such a case, no error will be logged on the list
and 'mca_count' will not be incremented.
Prior to this patch, all CPUs that received the exception would spin
forever.
With this change, the CPU that detects the error but finds the freelist
empty will proceed to panic the machine, ending the deadlock.
A follow-up to r260457.
Reported by: Ryan Libby <rlibby at gmail.com>
Reviewed by: jhb@
Sponsored by: Dell EMC Isilon
Differential Revision: https://reviews.freebsd.org/D10536
Fix end coordinate of the drawable area of border. Although the name tr_end
suggests it is the end coordinate, tr_end.tp_row is width and tr_end.tp_col
is height of the drawable area in reality.
Allow some control over enabled capabilities for if_vlan.
It improves interoperability with if_bridge, which may need to disable
some capabilities not supported by other members. IMHO there is still
open question about LRO capability, which may need to be disabled on
physical interface.
- For security reasons by default listen on localhost address,
not on wildcard. [1]
- Move the default port assignment from pci_fbuf.c to rfb.c,
to avoid polluting pci_fbuf.c with network things.
https://www.illumos.org/issues/7740
The problem is that dbuf_findbp will return ENOENT if the block it's
trying to find is beyond the end of the file. If that happens, we assume
there is no birth time, and so we lose that information when we write
out new blkptrs. We should teach dbuf_findbp to look for things that are
beyond the current end, but not beyond the absolute end of the file.
To verify, create a large file, truncate it to a short length, and then
write beyond the end. Check with zdb to make sure that there are no
holes with birth time zero (will appear as gaps).
Reviewed by: Steve Gonczi <steve.gonczi@delphix.com>
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Approved by: Dan McDonald <danmcd@omniti.com>
Author: Paul Dagnelie <pcd@delphix.com>
https://www.illumos.org/issues/7743
When loading a pool that had been created before the existance of
per-vdev zaps, on a system that knows about per-vdev zaps, the
per-vdev zaps will not be allocated and initialized.
This appears to be because the logic that would have done so, in
spa_sync_config_object(), is not reached under normal operation. It is
only reached if spa_config_dirty_list is non-empty.
The fix is to add another `AVZ_ACTION_` enum that will allow this code
to be reached when we detect that we're loading an old pool, even when
there are no dirty configs.
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Don Brady <don.brady@intel.com>
Approved by: Robert Mustacchi <rm@joyent.com>
Author: Paul Dagnelie <pcd@delphix.com>
https://www.illumos.org/issues/7613
metaslab_t:ms_freetree[TXG_SIZE] is only used in syncing context. We should
replace it with two trees: the freeing tree (ranges that we are freeing this
syncing txg) and the freed tree (ranges which have been freed this txg).
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Alex Reece <alex@delphix.com>
Approved by: Dan McDonald <danmcd@omniti.com>
Author: Matthew Ahrens <mahrens@delphix.com>
Add asserts to verify stability of struct proc and struct thread layouts.
Some notes:
- Only i386 and amd64 layouts are checked, other Tier-1 (or close to
it) architectures would benefit from the same check.
- Unconditional enabling of the asserts depend on the stability of locks
memory layout. If locks are optimized to avoid bloat when some debugging
or profiling features turned off, it makes sense to only assert layout
for production configs.
Reviewed by: badger, emaste, jhb, vangyzen
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
Differential revision: https://reviews.freebsd.org/D10526
https://www.illumos.org/issues/7586
The #ifdef __lint in dmu.h is ugly, and it would be nice not to duplicate it if
we add other inline functions into header files in ZFS, especially since it is
difficult to make any other solution work across all compilation targets. We
should switch to disabling the lint flags that are failing instead.
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Approved by: Dan McDonald <danmcd@omniti.com>
Author: Dan Kimmel <dan.kimmel@delphix.com>
armv8 has support for optional CRC32C instructions. This patch checks if they are
available and if that is true make use of them.
Thank you very much to Andrew Turner for providing help and review the patch!
Reviewed by: andrew
MFC after: 1 week
Differential Revision: https://reviews.freebsd.org/D10499
https://www.illumos.org/issues/7580
We need to prevent any reader whenever we're about the zero out all the
blkptrs. To do this we need to grab the dn_struct_rwlock as writer in
dbuf_write_children_ready and free_children just prior to calling bzero.
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Reviewed by: Steve Gonczi <steve.gonczi@delphix.com>
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Approved by: Dan McDonald <danmcd@omniti.com>
Author: George Wilson <george.wilson@delphix.com>
Various fixes for PCI _OSC handling so HotPlug works again.
- Rename the default implementation of 'pcib_request_feature' and add
a pcib_request_feature() wrapper function (as is often done for
new-bus APIs implemented via kobj) that accepts a single function.
Previously the call to pcib_request_feature() ended up invoking the
method on the great-great-grandparent of the bridge device instead
of the grandparent. For a bridge that was a direct child of pci0 on
x86 this resulted in the method skipping over the Host-PCI bridge
driver and being invoked against nexus0
- When invoking _OSC from a Host-PCI bridge driver, invoke
device_get_softc() against the Host-PCI bridge device instead of the
child bridge that is requesting HotPlug. Using the wrong softc data
resulted in garbage being passed for the ACPI handle causing the
_OSC call to fail.
- While here, perform some other cleanups to _OSC handling in the ACPI
Host-PCI bridge driver:
- Don't invoke _OSC when requesting a control that has already been
granted by the firmware.
- Don't set the first word of the capability array before invoking
_OSC. This word is always set explicitly by acpi_EvaluateOSC()
since it is UUID-independent.
- Don't modify the set of granted controls unless _OSC doesn't exist
(which is treated as always successful), or the _OSC method
doesn't fail.
- Don't require an _OSC status of 0 for success. _OSC always
returns the updated control mask even if it returns a non-zero
status in the first word.
- Whine if _OSC ever tries to revoke a previously-granted control.
(It is not supposed to do that.)
- While here, add constants for the _OSC status word in acpivar.h
(though currently unused).
Reported by: adrian
Reviewed by: imp
MFC after: 1 week
Tested on: Lenovo x220
Differential Revision: https://reviews.freebsd.org/D10520
https://www.illumos.org/issues/7606
When importing a pool with a large number of filesystems within the same
parent filesystem, we see that dmu_objset_find_dp() takes a long time.
It is called from 3 places: spa_check_logs(), spa_ld_claim_log_blocks(),
and spa_load_verify().
There are several ways to improve performance here:
1. We don't really need to do spa_check_logs() or
spa_ld_claim_log_blocks() if the pool was closed cleanly.
2. spa_load_verify() uses dmu_objset_find_dp() to check that no
datasets have too long of names.
3. dmu_objset_find_dp() is slow because it's doing
zap_value_search() (which is O(N sibling datasets)) to determine
the name of each dsl_dir when it's opened. In this case we
actually know the name when we are opening it, so we can provide
it and avoid the lookup.
This change implements fix #3 from the above list; i.e. make
dmu_objset_find_dp() provide the name of the dataset so that we don't
have to search for it.
Reviewed by: Steve Gonczi <steve.gonczi@delphix.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Prashanth Sreenivasa <prashksp@gmail.com>
Approved by: Gordon Ross <gordon.w.ross@gmail.com>
Author: Matthew Ahrens <mahrens@delphix.com>
Prefer to use real virtual address over direct map address in the
linux_page_address() function in the LinuxKPI. This solves an issue
where the return value from linux_page_address() is passed to
kmem_free().
Fix handling of a NFSv4.1 callback reply from the session cache.
The nfsv4_seqsession() call returns NFSERR_REPLYFROMCACHE when it has a
reply in the session, due to a requestor retry. The code erroneously
assumed a return of 0 for this case. This patch fixes this and adds
a KASSERT(). This would be an extremely rare occurrence. It was found
during code inspection during the pNFS server development.