are actually caused by a buf with both VNCLEAN and VNDIRTY set. In
the traces it is clear that the buf is removed from the dirty queue while
it is actually on the clean queue which leaves the tail pointer set.
Assert that both flags are not set in buf_vlist_add and buf_vlist_remove.
Sponsored by: Isilon Systems, Inc.
Approved by: re (blanket vfs)
cache_zap() to clear the v_dd pointers when a directory vnode is forcibly
discarded. For this to work, all vnodes with v_dd pointers to a directory
must also have name cache entries linked via v_cache_dst to that dvp
otherwise we could not find them at cache_purge() time. The following
code snipit could break this guarantee by unlinking a directory before
fetching it's dotdot. The dotdot lookup would initialize the v_dd field
of the unlinked directory which could never be cleared. To fix this
we don't initialize v_dd for orphaned vnodes.
printf("rmdir: %d\n", rmdir("../foo")); /* foo is cwd */
printf("chdir: %d\n", chdir(".."));
printf("%s\n", getwd(NULL));
Sponsored by: Isilon Systems, Inc.
Discovered by: kkenn
Approved by: re (blanket vfs)
not only means that it's possible (though unlikely) that we hand out
differing tags for the same bus space, it also means that the tags
we handed out are not used during bus enumeration. Both affect our
ability to compare tags. Fix the first by initializing our tags only
once. Fix the second by testing if one of the tags to compare is our
tag and the other is a busspace_isa_{io|mem} tag and declare them
equal if so.
This fixes using uart(4) as the serial console on a ds10. That is,
the low-level console worked, but we could not match the resources
to one of the UARTs found during bus enumeration, which prevented
uart(4) from becoming the console in single- or multi-user mode.
Approved by: re (kensmith)
MFC after: 2 days
Thanks to: all involved in getting a ds10 to me; directly or indirectly.
Special thanks to: Dave Knight, ISC (for not scratching my Porsche :-)
pending discussion of how implementation would proceed. Applications
like -lc_r expect select(3) to match the EAGAIN-status of IO
functions.
Approved by: re
- do not use static memory as we are under a shared lock only
- properly rtfree routes allocated with rtalloc
- rename to verify_path6()
- implement the full functionality of the IPv4 version
Also make O_ANTISPOOF work with IPv6.
Reviewed by: gnn
Approved by: re (blanket)
kernel mode, always use the curthread pmap instead. There are valid cases
were we can fault on a user address from the kernel without pcb_onfault
being set.
Approved by: re (blanket)
contaminated with the GPL code. While this information was present in
the COPYRIGHT.INFO file, it is FreeBSD's standard practice to, where
possible, include explicit license information in files.
Approved by: release engineer (scottl)
ref while we're calling vgone(). This prevents transient refs from
re-adding us to the free list. Previously, a vfree() triggered via
vinvalbuf() getting rid of all of a vnode's pages could place a partially
destructed vnode on the free list where vtryrecycle() could find it. The
first call to vtryrecycle would hang up on the vnode lock, but when it
failed it would place a now dead vnode onto the free list, and another
call to vtryrecycle() would free an already free vnode. There were many
complications of having a zero ref count while freeing which can now go
away.
- Change vdropl() to release the interlock before returning. All callers
now respect this, so vdropl() directly frees VI_DOOMED vnodes once the
last ref is dropped. This means that we'll never have VI_DOOMED vnodes
on the free list.
- Seperate v_incr_usecount() into v_incr_usecount(), v_decr_usecount() and
v_decr_useonly(). The incr/decr split is so that incr usecount can
return with the interlock still held while decr drops the interlock so
it can call vdropl() which will potentially free the vnode. The calling
function can't drop the lock of an already free'd node. v_decr_useonly()
drops a usecount without droping the hold count. This is done so the
usecount reaches zero in vput() before we recycle, however the holdcount
is still 1 which prevents any new references from placing the vnode
back on the free list.
- Fix vnlrureclaim() to vhold the vnode since it doesn't do a vget(). We
wouldn't want vnlrureclaim() to bump the usecount since this has
different semantics. Also change vnlrureclaim() to do a NOWAIT on the
vn_lock. When this function runs we're usually in a desperate situation
and we wouldn't want to wait for any specific vnode to be released.
- Fix a bunch of misc comments to reflect the new behavior.
- Add vhold() and vdrop() to vflush() for the same reasons that we do in
vlrureclaim(). Previously we held no reference and a vnode could have
been freed while we were waiting on the lock.
- Get rid of vlruvp() and vfreehead(). Neither are used. vlruvp() should
really be rethought before it's reintroduced.
- vgonel() always returns with the vnode locked now and never puts the
vnode back on a free list. The vnode will be freed as soon as the last
reference is released.
Sponsored by: Isilon Systems, Inc.
Debugging help from: Kris Kennaway, Peter Holm
Approved by: re (blanket vfs)
Hopefully this fixes ed(4) under qemu. I'm shocked that real hardware
is apparently working with these bugs.
Approved by: re (ifnet blanket)
Pointy hat: brooks
of the clean and dirty lists. This is in an attempt to catch the wrong
bufobj problem sooner.
- In vgonel() don't acquire an extra reference in the active case, the
vnode lock and VI_DOOMED protect us from recursively cleaning.
- Also in vgonel() clean up some stale comments.
Sponsored by: Isilon Systems, Inc.
Approved by: re (blanket vfs)
early. I've moved it all the way to the top rather than part way up as
the submitter did.
Submitted by: Jung-uk Kim <jkim at niksun dot com>
Reported by: submitter, le, dougb
Approved by: re (ifnet blanket)
function pointer to the vga render dispatch table and initialized it with
vga_nop. The problem is that vga_nop() is a varargs function, and the
table declares a non-varargs function pointer. On amd64 (and I think ppc),
mixing varargs and non-varargs function pointers is fatal.
Change vga_nop() and gfb_nop() from varargs to non-varargs do-nothing
functions. This stops the stack corruption that only happened on amd64.
Approved by: re (scottl)
used to ensure that we weren't exiting the syscall with a lock still
held. This wasn't safe, however, because we'd already executed a vput()
and on a loaded system the vnode may have been free'd by the time we
assert. This functionality is also handled by the td_locks assert in
userret, which doesn't tell you what the syscall was, but will at least
panic before you deadlock.
Sponsored by: Isilon Systems, Inc.
Discovred by: Peter Holm
Approved by: re (blanket vfs)
anyway and it's not used outside of vfs_subr.c.
- Change vgonel() to accept a parameter which determines whether or not
we'll put the vnode on the free list when we're done.
- Use the new vgonel() parameter rather than VI_DOOMED to signal our
intentions in vtryrecycle().
- In vgonel() return if VI_DOOMED is already set, this vnode has already
been reclaimed.
Sponsored by: Isilon Systems, Inc.
this is happening at the moment and sometimes causing panics later on the
package cluster when we bremfree() a buf whose delayed bremfree() did not
previously happen.
Sponsored by: Isilon Systems, Inc.
most of the code to deal with them has been dead for sometime. Simplify
the code by doing an insert sort hinted by the current head position.
Met with apathy by: arch@
on an IPv4 packet as these variables are uninitialized if not. This used to
allow arbitrary IPv6 packets depending on the value in the uninitialized
variables.
Some opcodes (most noteably O_REJECT) do not support IPv6 at all right now.
Reviewed by: brooks, glebius
Security: IPFW might pass IPv6 packets depending on stack contents.
Approved by: re (blanket)
I introduce a very small race here (some file system can be mounted or
unmounted between 'count' calculation and file systems list creation),
but it is harmless.
Found by: FreeBSD Kernel Stress Test Suite: http://www.holm.cc/stress/
Reported by: Peter Holm <peter@holm.cc>
fails.
Move detaching the ifnet from the ifindex_table into if_free so we can
both keep the sanity checks and actually delete the ifnets. [0]
Reported by: gallatin [0]
Approved by: re (blanket)
It can be used to panic the kernel by giving too big value.
Fix it by moving allocation and size verification into kern_getfsstat().
This even simplifies kern_getfsstat() consumers, but destroys symmetry -
memory is allocated inside kern_getfsstat(), but has to be freed by the
caller.
Found by: FreeBSD Kernel Stress Test Suite: http://www.holm.cc/stress/
Reported by: Peter Holm <peter@holm.cc>
o getsockopt(SO_ACCEPTFILTER) always returns success on listen socket
even we didn't install accept filter on the socket.
o Fix these bugs and add regression tests for them.
Submitted by: Igor Sysoev [1]
Reviewed by: alfred
MFC after: 2 weeks
to initialize the buffer array in ata_raid_attach() by removing the
initializer. There's no memset(?) in the kernel. Instead, assign
'\0' to the first element. The buffer array holds strings only, so
this is functionally equivalent.
Applies to: ia64
Tripped over by: tinderbox
events could be added to cover other interesting details.
- Add some VNASSERTs to discover places where we access vnodes after
they have been uma_zfree'd before we try to free them again.
- Add a few more VNASSERTs to vdestroy() to be certain that the vnode is
really unused.
Sponsored by: Isilon Systems, Inc.
in that KTR_VFS will be hand placed, while KTR_VOP traces the individual
vnode operations and is generated by vnode_if.awk.
- Add a comment describing KTR_VOP.
I missed. Since I did no rearrange any softcs, casting the result of
device_get_softc() to (struct ifnet **) and derefrencing it yeilds a
pointer to the ifp. This makes at least vr(4) nics work.
many regions checked again and again despite knowing the pages
contained were not usable and only satisfied the alignment constraints
This case was compounded, especially for large allocations, by the
practice of looping from the top of memory so as to keep out of the
important low-memory regions. While the old contigmalloc(9) has the
same problem, it is not as noticeable due to looping from the low
memory to high.
This degenerate case is fixed, as well as reversing the sense of the
rest of the loops within it, to provide a tremendous speed increase.
This makes the best case O(n * VM overhead) much more likely than the
worst case O(4 * VM overhead). For comparison, the worst case for old
contigmalloc would be O(5 * VM overhead) in addition to its strategy
of turning used memory into free being highly pessimal.
Also, fix a bug that in practice most likely couldn't have been triggered,
int the new contigmalloc(9): it walked backwards from the end of memory
without accounting for how many pages it needed. Potentially, nonexistant
pages could have been mapped. This hasn't occurred because the kernel
generally requests as its first contigmalloc(9) a single page.
Reported by: Nicolas Dehaine <nicko@stbernard.com>, wes
MFC After: 1 month
More testing by: Nicolas Dehaine <nicko@stbernard.com>, wes
atomic write request, it can fill the buffer cache with the entirety
of that write in order to handle retries. However, it never drops
the vnode lock, or else it wouldn't be atomic, so it ends up waiting
indefinitely for more buf memory that cannot be gotten as it has it
all, and it waits in an uncancellable state.
To fix this, hibufspace is exported and scaled to a reasonable
fraction. This is used as the limit of how much of an atomic write
request by the NFS client will be handled asynchronously. If the
request is larger than this, it will be turned into a synchronous
request which won't deadlock the system. It's possible this value is
far off from what is required by some, so it shall be tunable as soon
as mount_nfs(8) learns of the new field.
The slowdown between an asynchronous and a synchronous write on NFS
appears to be on the order of 2x-4x.
General nod by: gad
MFC after: 2 weeks
More testing: wes
PR: kern/79208
well worth the bloat.
- Change the formatting of 'show ktr' slightly to accommodate the
additional field. Remove a tab from the verbose output and place the
actual trace data after a : so it is more easy to understand which
part is the event and which is part of the record.
psm(4), ukbd(4), ums(4) and usb(4) on by default. Modulo some nits with
the most annoying one probably being USB keyboards no longer working at
the OFW boot prompt after halting FreeBSD these drivers work fine on
sparc64 including X and there's nothing left that I'd consider a show-
stopper. I.e. graphical consoles on sun4u machines should either work
out of the box or by plugging in a card that is supported by either
creator(4) or machfb(4). The exception obviously are SBus-only machines
without UPA slots like some Ultra 1 (but which also still lack support
in other areas) and certain Exx0 (but which probably are mainly used
with serial consoles anyway). I'll try to add a cgsix(4) for these later
as Sun CG6 cards are probably the most common SBus framebuffer cards in
sun4u machines. I however don't see much sense in adding drivers for the
dozen of SBus framebuffers that were destined for sparc v8 machines.
The rest of the USB drivers aren't enabled as I'm only aware of ukbd(4)
and ums(4) as well as ohci(4) working with the on-board ALI M5237 and
Sun PCIO-2 controllers. Aue(4) definitely doesn't work on sparc64, yet.
Thanks to:
- Jake for the initial work on syscons(4) on sparc64 and creator(4).
- Marcel for uart(4) and especially for its support for the SCCs which
are only used on sparc64 so far. In various regards it wouldn't have
been possible to enable syscons(4) by default on sparc64, yet, without
uart(4).
- All that tested patches.
Ok'ed by: scottl (RE hat), tmm
files after they were repo-copied to sys/dev/atkbdc. The sources of
atkbdc(4) and its children were moved to the new location in preparation
for adding an EBus front-end to atkbdc(4) for use on sparc64; i.e. in
order to not further scatter them over the whole tree which would have
been the result of adding atkbdc_ebus.c in e.g. sys/sparc64/ebus. Another
reason for the repo-copies was that some of the sources were misfiled,
e.g. sys/isa/atkbd_isa.c wasn't ISA-specific at all but for hanging
atkbd(4) off of atkbdc(4) and was renamed to atkbd_atkbdc.c accordingly.
Most of sys/isa/psm.c, i.e. expect for its PSMC PNP part, also isn't
ISA-specific.
- Separate the parts of atkbdc_isa.c which aren't actually ISA-specific
but are shareable between different atkbdc(4) bus front-ends into
atkbdc_subr.c (repo-copied from atkbdc_isa.c). While here use
bus_generic_rl_alloc_resource() and bus_generic_rl_release_resource()
respectively in atkbdc_isa.c instead of rolling own versions.
- Add sparc64 MD bits to atkbdc(4) and atkbd(4) and an EBus front-end for
atkbdc(4). PS/2 controllers and input devices are used on a couple of
Sun OEM boards and occur on either the EBus or the ISA bus. Depending on
the board it's either the only on-board mean to connect a keyboard and
mouse or an alternative to either RS232 or USB devices.
- Wrap the PSMC PNP part of psm.c in #ifdef DEV_ISA so it can be compiled
without isa(4) (e.g. for EBus-only machines). This ISA-specific part
isn't separated into its own source file, yet, as it requires more work
than was feasible for 6.0 in order to do it in a clean way. Actually
philip@ is working on a rewrite of psm(4) so a more comprehensive
clean-up and separation of hardware dependent and independent parts is
expected to happen after 6.0.
Tested on: i386, sparc64 (AX1105, AXe and AXi boards)
Reviewed by: philip
of just dropping the lock around the ip_output call. This used to cause
corrupted state tree walks for some call-paths.
In a second stage all callouts will be marked MPSAFE according to the
setting of mpsafenet.
Reported and tested by: Matthew Grooms <mgrooms at seton dot org>
MFC after: 3 days
X-MFC after: Marking callouts MPSAFE + 1 week
struct ifnet or the layer 2 common structure it was embedded in have
been replaced with a struct ifnet pointer to be filled by a call to the
new function, if_alloc(). The layer 2 common structure is also allocated
via if_alloc() based on the interface type. It is hung off the new
struct ifnet member, if_l2com.
This change removes the size of these structures from the kernel ABI and
will allow us to better manage them as interfaces come and go.
Other changes of note:
- Struct arpcom is no longer referenced in normal interface code.
Instead the Ethernet address is accessed via the IFP2ENADDR() macro.
To enforce this ac_enaddr has been renamed to _ac_enaddr.
- The second argument to ether_ifattach is now always the mac address
from driver private storage rather than sometimes being ac_enaddr.
Reviewed by: sobomax, sam
it; instead pass the space occupied by the header down into the
crypto modules (except in the demic case which needs it only when
doing int in s/w)
o while here fix defrag to strip the header from 2nd and later frames
o teach decap code how to handle 4-address frames
- Do not edit pullup_len outside M_CHECK macro.
- Do not reimplement NG_FWD_NEW_DATA().
- Remove redundant check for item being not NULL.
Submitted by: ru
do the subsequent ip_output() in IPFW. In ipfw_tick(), the keep-alive
packets must be generated from the data that resides under the
stateful lock, but they must not be sent at that time, as this would
cause a lock order reversal with the normal ordering (interface's
lock, then locks belonging to the pfil hooks).
In practice, this caused deadlocks when using IPFW and if_bridge(4)
together to do stateful transparent filtering.
MFC after: 1 week
- Initialize val_ec with the content of the volume EC register
for ACPI_IBM_METHOD_VOLUME and ACPI_IBM_METHOD_MUTE in acpi_ibm_sysctl_set()
if there is no CMOS handle present. This fixes setting volume and mute on
such models.
Submitted by: ru
Approved by: philip
hack MSS of packets outgoing via interface with small MTU, to workaround
path MTU discovery problems.
Written by Alexey Popov, with some cleanups from me. There are also plans
to improve mpd port, so that it uses this node, instead of doing MSS
hacking in userland, when 'enable tcpmssfix' option is on.
Submitted by: Alexey Popov <lollypop@flexuser.ru>
vm_page's machine-dependent fields. Use this function in
vm_pageq_add_new_page() so that the vm_page's machine-dependent and
machine-independent fields are initialized at the same time.
Remove code from pmap_init() for initializing the vm_page's
machine-dependent fields.
Remove stale comments from pmap_init().
Eliminate the Boolean variable pmap_initialized from the alpha, amd64,
i386, and ia64 pmap implementations. Its use is no longer required
because of the above changes and earlier changes that result in physical
memory that is being mapped at initialization time being mapped without
pv entries.
Tested by: cognet, kensmith, marcel
UFS by:
- Making the pre and post hooks for the VOP functions work even when
DEBUG_VFS_LOCKS is not defined.
- Moving the KNOTE activations into the corresponding VOP hooks.
- Creating a MNTK_NOKNOTE flag for the mnt_kern_flag field of struct
mount that permits filesystems to disable the new behavior.
- Creating a default VOP_KQFILTER function: vfs_kqfilter()
My benchmarks have not revealed any performance degradation.
Reviewed by: jeff, bde
Approved by: rwatson, jmg (kqueue changes), grehan (mentor)
- Restructured for easier extensibility and maintainability
- To be more uniform with the other ACPI extras drivers and to better reflect
their actual meaning, some sysctls were moved:
o brightness -> lcd_brightness
o keylight -> thinklight
o enable -> events
o misckey -> hotkey
o avail_mask -> availmask
o key_mask -> eventmask
- New "initialmask" sysctl, which holds the initial eventmask
- The "wlan" sysctl is now read-only, since writing to it didn't have
any effect
- The "version" sysctl was removed, since it seems to be the same (0x100)
on all models I have seen
- Support for more hotkeys by the "hotkey" sysctl
- Improved support of ACPI events. Disabled by default, since it unexpectedly
changes the behaviour of some keys. (on my T41p there are now 24 different
keypress events that get reported)
- write support for: volume, mute, lcd_brightness and thinklight
- led(4) interface for the thinklight [1]
- New sysctls "fan" and "fan_speed" to support reading of fan status and speed
- New sysctl "thermal" to support reading of up to 8 thermal sensors
Reviewed by: philip
Approved by: philip
Submitted by: simon [1]
Inspired by: The Linux ibm_acpi driver by Borislav Deianov
http://ibm-acpi.sourceforge.net/
The ThinkPad Button program (tpb) by Markus Braun
http://www.nongnu.org/tpb/
Thanks to: brueffer, dvl, njl, philip, simon, takawata and the many
testers from freebsd-acpi@ and freebsd-mobile@
- Implement sampling modes and logging support in hwpmc(4).
- Separate MI and MD parts of hwpmc(4) and allow sharing of
PMC implementations across different architectures.
Add support for P4 (EMT64) style PMCs to the amd64 code.
- New pmcstat(8) options: -E (exit time counts) -W (counts
every context switch), -R (print log file).
- pmc(3) API changes, improve our ability to keep ABI compatibility
in the future. Add more 'alias' names for commonly used events.
- bug fixes & documentation.
makes the amount of spare room consistent across architectures, and gets
rid of some cases where space was wasted-and-unusable because types have
different sizes & alignment on different hardware platforms. This change
causes more variables to move around on i386, and is much less disruptive
on other platforms. This has been tested on i386, ppc, and sparc, and has
at least been compiled on everything but arm.
Reviewed by: no objections from freebsd-arch, bde
and extend its functionality:
value policy
0 show all mount-points without any restrictions
1 show only mount-points below jail's chroot and show only part of the
mount-point's path (if jail's chroot directory is /jails/foo and
mount-point is /jails/foo/usr/home only /usr/home will be shown)
2 show only mount-point where jail's chroot directory is placed.
Default value is 2.
Discussed with: rwatson
security.bsd.see_other_uids is set to 0, etc.
One can check if invisible process is active, by doing:
# ktrace -p <pid>
If ktrace returns 'Operation not permitted' the process is alive and
if returns 'No such process' there is no such process.
MFC after: 1 week
