If we are at cancellation point, always work as deferred mode despite
whether asynchronous mode is turned on or not, this always gives us a
chance to decide whether thread should be canceled or not in
cancellation points.
Support new variable NANO_LABEL. When set to a non-null string,
nanobsd will build a system that uses this label (via
/dev/ufs/${NANO_LABEL}sX) in preference to NANO_DRIVE (well, it forces
NANO_DRIVE to be ufs/${NANO_LABEL}). This allows images that will
boot off usb stick or CF card easily well.
There is no change if you don't set this variable.
brian [Mon, 20 Sep 2010 22:04:59 +0000 (22:04 +0000)]
Add a geli resize subcommand to resize encrypted filesystems prior
to growing the filesystem.
Refuse to attach providers where the metadata provider size is
wrong. This makes post-boot attaches behave consistently with
pre-boot attaches. Also refuse to restore metadata to a provider
of the wrong size without the new -f switch. The new -f switch
forces the metadata restoration despite the provider size, and
updates the provider size in the restored metadata to the correct
value.
First step at adopting FreeBSD to support PSARC/2010/029. This makes
acl_is_trivial_np(3) properly recognize the new trivial ACLs. From
the user point of view, that means "ls -l" no longer shows plus signs
for all the files when running ZFS v28.
Tweak the stats exported by the e1000 drivers:
- Add a single sysctl procedure to all three drivers to read an arbitrary
register (the register is passed as arg2). Use it to replace existing
routines in igb(4) that used a separate routine for each register, and
to add support for missing stats in em(4) and lem(4).
- Move the 'rx_overruns' and 'watchdog_timeouts' stats out of the MAC stats
section as they are driver stats, not MAC counters.
- Simplify the code that creates per-queue stats in igb(4) to use a single
loop and remove duplicated code.
- Properly read all 64 bits of the 'good octets received/transmitted' in
em(4) and lem(4).
- Actually read the interrupt count registers in em(4), and drop the
'host to card' sysctl stats from em(4) as they are not implemented in
any of the hardware this driver supports.
- Restore several stats to em(4) that were lost in the earlier stats
conversion including per-queue stats.
- Export several MAC stats in em(4) that were exported in igb(4) but not
in em(4).
- Export stats in lem(4) using individual sysctls as in em(4) and igb(4).
- 64 bit fixes for ifnlge.c
- Use m_nextpkt to save mbuf vaddr on 64 bit, we cannot store the
64 bit vaddr in the 40bit freeback field.
- remove unused code and unnecessary variables.
- use xlr_io_mmio macro instead of adding io base address
- rewrite GPIO related code to fixup nlge using xlr_write_reg and DELAY
- support for engg boards major num 11 and 12
- add xlr_paddr_lw() to load 32bit value from physical address, fix
inline assembly
- style fixes
Unbreak the build on strong-aligned architectures (arm, ia64).
Casting from (char *) to (struct ufs1_dinode *) changes the
alignment requirement of the pointer and GCC does not know that
the pointer is adequately aligned (due to malloc(3)), and warns
about it. Cast to (void *) first to by-pass the check.
Allow a POSIX shared memory object that is opened for read but not for
write to nonetheless be mapped PROT_WRITE and MAP_PRIVATE, i.e.,
copy-on-write.
(This is a regression in the new implementation of POSIX shared memory
objects that is used by HEAD and RELENG_8. This bug does not exist in
RELENG_7's user-level, file-based implementation.)
Make refinements to r212824. In particular, don't make
vm_map_unlock_nodefer() part of the synchronization interface for maps.
Add comments to vm_map_unlock_and_wait() and vm_map_wakeup() describing
how they should be used. In particular, describe the deferred deallocations
issue with vm_map_unlock_and_wait().
Redo the implementation of vm_map_unlock_and_wait() so that it passes
along the caller's file and line information, just like the other map
locking primitives.
ed [Sun, 19 Sep 2010 16:35:42 +0000 (16:35 +0000)]
Just make callout devices and /dev/console force CLOCAL on open().
Instead of adding custom checks to wait for DCD on open(), just modify
the termios structure to set CLOCAL. This means SIGHUP is no longer
generated when losing DCD as well.
Rewrite parts of the calibration code which is run while bringing up
the device:
- unobscure some of the code by moving it into its own functions
- get rid of some magic numbers
- create similar structure as the reference driver has, this should
make further syncs easier
Add preliminary support for the Lanner MR-955. It boots multi-user but there
seem to be problems both with the on-board Ethernet interfaces and the em(4)
interfaces on PCI under FreeBSD.
Thanks to Lanner for providing access to hardware.
brian [Sun, 19 Sep 2010 08:18:56 +0000 (08:18 +0000)]
Revise r197763 which fixes filesystem corruption when extending
into un-zeroed storage.
The original patch was questioned by Kirk as it forces the filesystem
to do excessive work initialising inodes on first use, and was never
MFC'd. This change mimics the newfs(8) approach of zeroing two
blocks of inodes for each new cylinder group.
- _Unwind_Resume function is not used, remove it.
- Use a store barrier to make sure uwl_forcedunwind is lastest thing
other threads can see.
- Add some comments.
Fix nfsrv_freeallnfslocks() in the experimental NFSv4 server so that
it frees local locks correctly upon close. In order for
nfsrv_localunlock() to work correctly, the lock can no longer be in
the lockowner's stateid list. As such, nfsrv_freenfslock() has to
be called before nfsrv_localunlock(), to get rid of the lock structure
on the lockowner's stateid list. This only affected operation when
local locks (vfs.newnfs.enable_locallocks=1) are enabled, which is
not the default at this time.
Fix the experimental NFSv4 server so that it performs local VOP_ADVLOCK()
unlock operations correctly. It was passing in F_SETLK instead of
F_UNLCK as the operation for the unlock case. This only affected
operation when local locking (vfs.newnfs.enable_locallocks=1) was enabled.
- For consistency, remove "." and ".." entries from de_dlist before
calling devfs_delete() (and thus possibly dropping dm_lock) in
devfs_rmdir_empty().
- Assert that we don't return doomed entries from devfs_find(). [1]
Adopt the deferring of object deallocation for the deleted map entries
on map unlock to the lock downgrade and later read unlock operation.
System map entries cannot be backed by OBJT_VNODE objects, no need to
defer deallocation for them. Map entries from user maps do not require
the owner map for deallocation, and can be accumulated in the
thread-local list for freeing when a user map is unlocked.
Move the collection of entries for deferred reclamation into
vm_map_delete(). Create helper vm_map_process_deferred(), that is
called from locations where processing is feasible. Do not process
deferred entries in vm_map_unlock_and_wait() since map_sleep_mtx is
held.
With reworking of the socket life cycle in 7.x, the need for a "sotryfree()"
was eliminated: all references to sockets are explicitly managed by sorele()
and the protocols. As such, garbage collect sotryfree(), and update
sofree() comments to make the new world order more clear.
MFC after: 3 days
Reported by: Anuranjan Shukla <anshukla at juniper dot net>
kern.sched.topology_spec sysctl: use step of 1 for group levels numeration
This is just a cosmetic change for prettier output.
'indent' variable/parameter serves two purposes: it specifies whitespace
indentation level and also implies cpu group level/depth.
It would have been better to split those two uses,
but for now just a simple change.
Restore pre-r212778 optimization, skipping timer reprogramming when it is
not neccessary. It allows to avoid time counter jump of up to 1/18s, when
base frequency slightly tuned via machdep.i8254_freq sysctl.
Fix few style things.
When global timer used at SMP system, update nextevent field on BSP before
sending IPI to other CPUs. Otherwise, other CPUs will try to honor stale
value, programming timer for zero interval. If timer is fast enough,
it caused extra interrupt before timer correctly reprogrammed by BSP.
Rework r210248. Although it fixed most of problems, it did not fix one
particular edge case where X-axis resolution is not multiple of font width.
Now we just advance enough scan lines, then deduct a partial scan line.
It is more intuitive than the previous code. Apply the same wisdom to EGA
and VGA planar renderers for consistency.
Reported by: David DEMELIER (demelier dot david at gmail dot com)
Before VirtualBox is fixed, mark with #ifdef what has to be done to make
it possible to boot from ZFS RAIDZ for example from within VirtualBox.
The problem with VirtualBox is that its BIOS reports only one disk present.
If we choose to ignore this report, we can find all the disks available.
We can't have this work-around to be turned on by default, because some broken
BIOSes report true when it comes to number of disks, but present the same disk
multiple times.
andre [Fri, 17 Sep 2010 22:05:27 +0000 (22:05 +0000)]
Rearrange the TSO code to make it more readable and to clearly
separate the decision logic, of whether we can do TSO, and the
calculation of the burst length into two distinct parts.
Change the way the TSO burst length calculation is done. While
TSO could do bursts of 65535 bytes that can't be represented in
ip_len together with the IP and TCP header. Account for that and
use IP_MAXPACKET instead of TCP_MAXWIN as base constant (both
have the same value of 64K). When more data is available prevent
less than MSS sized segments from being sent during the current
TSO burst.
Add two more KASSERTs to ensure the integrity of the packets.
Tested by: Ben Wilber <ben-at-desync com>
MFC after: 10 days
ken [Fri, 17 Sep 2010 21:53:56 +0000 (21:53 +0000)]
Fix a couple of mps problems.
When the driver is completely saturated with commands (1024 in the
case of the SAS2008 in my test system), I/O stops. If we tell CAM
that we have one less command slot than we have actually allocated,
everything works fine. We also need a few extra command slots to
allow for aborts and other task management commands to be sent down.
This needs more investigation to determine the root cause, but for
now this fixes things in my testing.
mps.c: Change a printf() to mps_printf().
mps_sas.c: Subtract 5 command slots when we tell CAM how many
commands we can handle.
Add some commented-out logic to print the contents
the CDBs for timed-out commands. This can help
in debugging devices that are timing out. This
will be uncommented once I bring some CAM changes in.
Reported by: Andrew Boyer <aboyer at averesystems dot com>
- Process some tx done messages in the transmit path, to ensure that
the XLR NA tx done FIFO does not overflow.
- Add a message ring handler API to process atmost a given number of
messages from a specified bucket mask. This will be used to process
the tx done messages
- Add a callout to restart transmit in the case transmit gets blocked.
- Update enable_msgring_int() and disable_msgring_int(), remove unused
args and make static.
Obtained from: Sriram Gorti (srgorti at netlogicmicro dot com)
amd64: reduce VM_KMEM_SIZE_SCALE to 1 allowing kernel to use more memory
KVA space is abundant on amd64, so there is no reason to limit kernel
map size to a fraction of available physical memory. In fact, it could
be larger than physical memory.
This should help with memory auto-tuning for ZFS and shouldn't affect
other workloads.
This should reduce number of circumstances for "kmem_map too small"
panics, but probably won't eliminate them entirely due to potential kmem
fragmentation.
In fact, you might want/need to limit maximum ARC size after this commit
if you need to resrve more memory for applications.
This change was discussed on arch@ and nobody said "don't do it".
zfs arc_reclaim_needed: remove redundant checks for arc_c_max and arc_c_max
Those checks are not present in upstream code and they are enforced in
actual calculations of delta by which ARC size can be grown or should be
reduced.
zfs arc_reclaim_needed: more reasonable threshold for available pages
vm_paging_target() is not a trigger of any kind for pageademon, but
rather a "soft" target for it when it's already triggered.
Thus, trying to keep 2048 pages above that level at the expense of ARC
was simply driving ARC size into the ground even with normal memory
loads.
Instead, use a threshold at which a pagedaemon scan is triggered, so
that ARC reclaiming helps with pagedaemon's task, but the latter still
recycles active and inactive pages.
Add one-shot mode support to attimer (i8254) event timer.
Unluckily, using one-shot mode is impossible, when same hardware used for
time counting. Introduce new tunable hint.attimer.0.timecounter, setting
which to 0 disables i8254 time counter and allows one-shot mode. Note,
that on some systems there may be no other reliable enough time counters,
so this tunable should be used with understanding.
ken [Thu, 16 Sep 2010 22:54:56 +0000 (22:54 +0000)]
MFp4 (//depot/projects/mps/...):
According to the MPT2 spec, task management commands are
serialized, and so no I/O should start while task management
commands are active.
So, to comply with that, freeze the SIM queue before we send any
task management commands (abort, target reset, etc.) down to the
IOC. We unfreeze the queue once the task management command
completes.
It isn't clear from the spec whether multiple simultaneous task
management commands are supported. Right now it is possible to
have multiple outstanding task management commands, especially in
the abort case. Multiple outstanding aborts do complete
successfully, so it may be supported.
We also don't yet have any recovery mechanism (e.g. reset the IOC)
if the task management command fails.
Increase the hostname display width. Many machines are named
with >12 characters. This still allows for 80 column output
with load averages of 99.99, 99.99, 99.99.
+ Add the SCRIPT environmental variable to the sub-shell. Its value is
the name of the typescript file.
+ Add the 'command' argument (if supplied on the command line) to the
typescript file. This creates a more complete typescript when invoked
this way - more equal to invoking script without supplying the 'command'
argument.
dim [Thu, 16 Sep 2010 21:14:28 +0000 (21:14 +0000)]
Make vendor/clang/dist exactly the same as upstream's r108428. Some
files and directories were already added/removed in the upstream
repository, but were not added/removed here, when the previous snapshot
was imported.
Note: I did not import the file test/Lexer/conflict-marker.c, because it
contains merge conflict markers on purpose, which upsets our pre-commit
hooks.
andre [Thu, 16 Sep 2010 21:06:45 +0000 (21:06 +0000)]
Remove the TCP inflight bandwidth limiter as announced in r211315
to give way for the pluggable congestion control framework. It is
the task of the congestion control algorithm to set the congestion
window and amount of inflight data without external interference.
In 'struct tcpcb' the variables previously used by the inflight
limiter are renamed to spares to keep the ABI intact and to have
some more space for future extensions.
In 'struct tcp_info' the variable 'tcpi_snd_bwnd' is not removed to
preserve the ABI. It is always set to 0.
In siftr.c in 'struct pkt_node' the variable 'snd_bwnd' is not removed
to preserve the ABI. It is always set to 0.
These unused variable in the various structures may be reused in the
future or garbage collected before the next release or at some other
point when an ABI change happens anyway for other reasons.
No MFC is planned. The inflight bandwidth limiter stays disabled by
default in the other branches but remains available.
- Compile fixes for 9.0, the previous version of this driver was
for FreeBSD 6.
- Add virtual address field in OperationDescriptor_t, we cannot use
MIPS_PHYS_TO_KSEG0 on physical address.
- Fixes for new message ring API
- Remove unused sys/mips/rmi/dev/sec/stats.h
- Whitespace fixes