https://www.illumos.org/issues/6569
The core issue I've found is that there is no throttle for how many
deletes get assigned to one TXG. As a results when deleting large files
we end up filling consecutive TXGs with deletes/frees, then write
throttling other (more important) ops.
There is an easy test case for this problem. Try deleting several
large files (at least 1/2 TB) while you do write ops on the same
pool. What we've seen is performance of these write ops (let's
call it sideload I/O) would drop to zero.
More specifically the problem is that dmu_free_long_range_impl()
can/will fill up all of the dirty data in the pool "instantly",
before many of the sideload ops can get in. So sideload
performance will be impacted until all the files are freed.
The solution we have tested at Nexenta (with positive results)
creates a relatively simple throttle for how many "free" ops we let
into one TXG.
However this solution exposes other problems that should also be
addressed. If we are to slow down freeing of data that means one
has to wait even longer (assuming vnode ref count of 1) to get shell
back after an rm or for NFS thread to finish the free-ing op.
To avoid this the proposed solution is to call zfs_inactive() async
for "large" files. Async freeing then begs for the reclaimed space
to be accounted for in the zpool's "freeing" prop.
The other issue with having a longer delete is the inability to
export/unmount for a longer period of time. The proposed solution
is to interrupt freeing of blocks when a fs is unmounted.
Author: Alek Pinchuk <alek@nexenta.com>
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Reviewed by: Sanjay Nadkarni <sanjay.nadkarni@nexenta.com>
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Approved by: Dan McDonald <danmcd@omniti.com>
Andriy Gapon [Fri, 20 Jan 2017 13:39:07 +0000 (13:39 +0000)]
don't abort writing of a core dump after EFAULT
It's possible to get EFAULT when writing a segment backed by a file
if the segment extends beyond the file.
The core dump could still be useful if we skip the rest of the segment
and proceed to other segements.
The skipped segment (or a portion of it) will be zero-filled.
While there, use 'const' to signify that core_write() only reads the
buffer and use __DECONST before calling vn_rdwr_inchunks() because it
can be used for both reading and writing.
Before the change:
kernel: Failed to write core file for process mmap_trunc_core (error 14)
kernel: pid 77718 (mmap_trunc_core), uid 1001: exited on signal 6
After the change:
kernel: Failed to fully fault in a core file segment at VA 0x800645000 with size 0x4000 to be written at offset 0x29000 for process mmap_trunc_core
kernel: pid 4901 (mmap_trunc_core), uid 1001: exited on signal 6 (core dumped)
Reviewed by: julian, kib
Obtained from: Panzura (older version of the change)
MFC after: 5 days
Sponsored by: Panzura
Differential Revision: https://reviews.freebsd.org/D9233
Andriy Gapon [Fri, 20 Jan 2017 13:21:27 +0000 (13:21 +0000)]
vmm_dev: work around a bogus error with gcc 6.3.0
The error is:
vmm_dev.c: In function 'alloc_memseg':
vmm_dev.c:261:11: error: null argument where non-null required (argument 1) [-Werror=nonnull]
Apparently, the gcc is unable to figure out that if a ternary operator
produced a non-NULL value once, then the operator with exactly the same
operands would produce the same value again.
Add runtime support for modifying the SQ and RQ completion event
moderation mode. The presence of this feature is indicated through the
firmware capabilities.
Enji Cooper [Fri, 20 Jan 2017 03:23:24 +0000 (03:23 +0000)]
Replace dot-dot relative pathing with SRCTOP-relative paths where possible
This reduces build output, need for recalculating paths, and makes it clearer
which paths are relative to what areas in the source tree. The change in
performance over a locally mounted UFS filesystem was negligible in my testing,
but this may more positively impact other filesystems like NFS.
LIBC_SRCTOP was left alone so Juniper (and other users) can continue to
manipulate lib/libc/Makefile (and other Makefile.inc's under lib/libc) as
include Makefiles with custom options.
Ed Maste [Fri, 20 Jan 2017 02:09:59 +0000 (02:09 +0000)]
libc: remove reference to nonexistent lib/locale directory
As far as I can tell this was introduced in r72406 and updated in several
subsequent revisions, but the lib/locale directory it referenced never
existed.
Reviewed by: ngie
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D9252
Pedro F. Giffuni [Fri, 20 Jan 2017 00:02:11 +0000 (00:02 +0000)]
mppc - Finish pluging NETGRAPH_MPPC_COMPRESSION.
There were several places where reference to compression were left
unfinished. Furthermore, KASSERTs contained references to MPPC_INVALID
which is not defined in the tree and therefore were sure to break with
INVARIANTS: comment them out.
Reported by: Eugene Grosbein
PR: 216265
MFC after: 3 days
Scott Long [Thu, 19 Jan 2017 21:47:50 +0000 (21:47 +0000)]
Rework the debug print API. Event printing no longer gets special handling.
All of the printing from the tables file now has wrappers so that the
handling is cleaner and it's possible to print something out (say, during
development) without having to fight the global debug flags. This re-org
will also make it easier to have the tables be compiled out at build time
if desired.
Other than fixing some minor bugs, there are no user-visible changes from
this change
Sponsored by: Netflix, Inc.
Differential Revision: D9238
https://www.illumos.org/issues/6569
The core issue I've found is that there is no throttle for how many
deletes get assigned to one TXG. As a results when deleting large files
we end up filling consecutive TXGs with deletes/frees, then write
throttling other (more important) ops.
There is an easy test case for this problem. Try deleting several
large files (at least 1/2 TB) while you do write ops on the same
pool. What we've seen is performance of these write ops (let's
call it sideload I/O) would drop to zero.
More specifically the problem is that dmu_free_long_range_impl()
can/will fill up all of the dirty data in the pool "instantly",
before many of the sideload ops can get in. So sideload
performance will be impacted until all the files are freed.
The solution we have tested at Nexenta (with positive results)
creates a relatively simple throttle for how many "free" ops we let
into one TXG.
However this solution exposes other problems that should also be
addressed. If we are to slow down freeing of data that means one
has to wait even longer (assuming vnode ref count of 1) to get shell
back after an rm or for NFS thread to finish the free-ing op.
To avoid this the proposed solution is to call zfs_inactive() async
for "large" files. Async freeing then begs for the reclaimed space
to be accounted for in the zpool's "freeing" prop.
The other issue with having a longer delete is the inability to
export/unmount for a longer period of time. The proposed solution
is to interrupt freeing of blocks when a fs is unmounted.
Author: Alek Pinchuk <alek@nexenta.com>
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Reviewed by: Sanjay Nadkarni <sanjay.nadkarni@nexenta.com>
Reviewed by: Pavel Zakharov <pavel.zakharov@delphix.com>
Approved by: Dan McDonald <danmcd@omniti.com>
Add mount option for tmpfs(5) to not use namecache.
The option "nonc" disables using of namecache for the created mount,
by default namecache is used. The rationale for the option is that
namecache duplicates the information which is already kept in memory
by tmpfs. Since it believed that namecache scales better than tmpfs,
or will scale better, do not enable the option by default. On the
other hand, smaller machines may benefit from lesser namecache
pressure.
Discussed with: mjg
Tested by: pho (as part of larger patch)
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
For directories, node->tn_spec.tn_dir.tn_parent pointer to the parent
is used. For non-directories, the implementation is naive, all
directory nodes are scanned to find a dirent linking the specified
node. This can be significantly improved by maintaining tn_parent for
all nodes, later.
Tested by: pho (as part of larger patch)
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
On dotdot lookup and fhtovp operations, it is possible for the file
represented by tmpfs node to be removed after the thread calculated
the pointer. In this case, tmpfs_alloc_vp() accesses freed memory.
Introduce the reference count on the nodes. The allnodes list from
tmpfs mount owns 1 reference, and threads performing unlocked
operations on the node, add one transient reference. Similarly, since
struct tmpfs_mount maintains the list where nodes are enlisted,
refcount it by one reference from struct mount and one reference from
each node on the list. Both nodes and tmpfs_mounts are removed when
refcount goes to zero.
Note that this means that nodes and tmpfs_mounts might survive some
time after the node is deleted or tmpfs_unmount() finished. The
tmpfs_alloc_vp() in these cases returns error either due to node
removal (tn_nlinks == 0) or because of insmntque1(9) error.
Tested by: pho (as part of larger patch)
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
Andriy Gapon [Thu, 19 Jan 2017 18:46:41 +0000 (18:46 +0000)]
fix a thread preemption regression in schedulers introduced in r270423
Commit r270423 fixed a regression in sched_yield() that was introduced
in earlier changes. Unfortunately, at the same time it introduced an
new regression. The problem is that SWT_RELINQUISH (6), like all other
SWT_* constants and unlike SW_* flags, is not a bit flag. So, (flags &
SWT_RELINQUISH) is true in cases where that was not really indended,
for example, with SWT_OWEPREEMPT (2) and SWT_REMOTEPREEMPT (11).
A straight forward fix would be to use (flags & SW_TYPE_MASK) ==
SWT_RELINQUISH, but my impression is that the switch types are designed
mostly for gathering statistics, not for influencing scheduling
decisions.
So, I decided that it would be better to check for SW_PREEMPT flag
instead. That's also the same flag that was checked before r239157.
I double-checked how that flag is used and I am confident that the flag
is set only in the places where we really have the preemption:
- critical_exit + td_owepreempt
- sched_preempt in the ULE scheduler
- sched_preempt in the 4BSD scheduler
Fix problem with suspend and resume when using Skylake chipsets. Make
sure the XHCI controller is reset after halting it. The problem is
clearly a BIOS bug as the suspend and resume is failing without
loading the XHCI driver. The same happens when using Linux and the
XHCI driver is not loaded.
Conrad Meyer [Thu, 19 Jan 2017 16:46:05 +0000 (16:46 +0000)]
ffs_vnops: Simplify extattr access
As suggested in r167010, use the structure type and macros to access and
modify UFS2 extended attributes. Add assertions that pointers are
aligned in places where we now access the data through a structure
pointer, instead of character-by-character.