mechanism so that early processing on mbufs can be performed before
a context switch to the NFS server threads. Because of this, if
the socket code is running without Giant, the NFS server also needs
to be able to run the upcall code without relying on the presence on
Giant. This change modifies the NFS server to run using a "giant
code lock" covering operation of the whole subsystem. Work is in
progress to move to data-based locking as part of the NFSv4 server
changes.
Introduce an NFS server subsystem lock, 'nfsd_mtx', and a set of
macros to operate on the lock:
NFSD_LOCK_ASSERT() Assert nfsd_mtx owned by current thread
NFSD_UNLOCK_ASSERT() Assert nfsd_mtx not owned by current thread
NFSD_LOCK_DONTCARE() Advisory: this function doesn't care
NFSD_LOCK() Lock nfsd_mtx
NFSD_UNLOCK() Unlock nfsd_mtx
Constify a number of global variables/structures in the NFS server
code, as they are not modified and contain constants only:
nfsrvv2_procid nfsrv_nfsv3_procid nonidempotent
nfsv2_repstat nfsv2_type nfsrv_nfsv3_procid
nfsrvv2_procid nfsrv_v2errmap nfsv3err_null
nfsv3err_getattr nfsv3err_setattr nfsv3err_lookup
nfsv3err_access nfsv3err_readlink nfsv3err_read
nfsv3err_write nfsv3err_create nfsv3err_mkdir
nfsv3err_symlink nfsv3err_mknod nfsv3err_remove
nfsv3err_rmdir nfsv3err_rename nfsv3err_link
nfsv3err_readdir nfsv3err_readdirplus nfsv3err_fsstat
nfsv3err_fsinfo nfsv3err_pathconf nfsv3err_commit
nfsrv_v3errmap
There are additional structures that should be constified but due
to their being passed into general purpose functions without const
arguments, I have not yet converted.
In general, acquire nfsd_mtx when accessing any of the global NFS
structures, including struct nfssvc_sock, struct nfsd, struct
nfsrv_descript.
Release nfsd_mtx whenever calling into VFS, and acquire Giant for
calls into VFS. Giant is not required for any part of the
operation of the NFS server with the exception of calls into VFS.
Giant will never by acquired in the upcall code path. However, it
may operate entirely covered by Giant, or not. If debug.mpsafenet
is set to 0, the system calls will acquire Giant across all
operations, and the upcall will assert Giant. As such, by default,
this enables locking and allows us to test assertions, but should not
cause any substantial new amount of code to be run without Giant.
Bugs should manifest in the form of lock assertion failures for now.
This approach is similar (but not identical) to modifications to the
BSD/OS NFS server code snapshot provided by BSDi as part of their
SMPng snapshot. The strategy is almost the same (single lock over
the NFS server), but differs in the following ways:
- Our NFS client and server code bases don't overlap, which means
both fewer bugs and easier locking (thanks Peter!). Also means
NFSD_*() as opposed to NFS_*().
- We make broad use of assertions, whereas the BSD/OS code does not.
- Made slightly different choices about how to handle macros building
packets but operating with side effects.
- We acquire Giant only when entering VFS from the NFS server daemon
threads.
- Serious bugs in BSD/OS implementation corrected -- the snapshot we
received was clearly a work in progress.
Based on ideas from: BSDi SMPng Snapshot
Reviewed by: rick@snowhite.cis.uoguelph.ca
Extensive testing by: kris
eg:
[foo]
...
matches any executable 'foo'
[/usr/bin/foo/]
...
matches any executable under the directory /usr/bin/foo/
Exact matches continue to function as before.
PR: bin/66769
Submitted-by: Dan Nelson
This change is possible since all the relevant drivers have been
fixed to set if_capenable properly. The field if_capabilities tracks
supported capabilities, which may be disabled administratively.
Inheriting checksum offload support from the parent interface isn't
that easy because the checksumming capabilities of the parent may be
toggled on the fly. Disable the code for now.
Kernel profiling for amd64's (normal and high resolution) should now
compile and work as (un)well as on i386's. It works better than user
profiling because:
- it uses _cyg_profile_func_*() instead of .mcount(), so it doesn't suffer
from gcc misspelling .mcount as mcount.
- it doesn't neglect saving %rax in .mcount().
The SMP case hasn't been tested. The high resolution subcase of this uses
the i8254, and as on i386's, the locking for this is deficient and the
i8254 is too inefficient. The acpi timer is also too inefficient.
- perfmon headers must be avoided until perfmon is supported.
- all call-used registers including return registers must be preserved
by .mcount(), etc., not quite as in profile.h. __cyg_profile_func_*()
don't require this, but they are (mis)implemented as aliases for
.mcount(), etc. so they preserve the registers.
- i386 ifdefs related to perfmon have not been adjusted yet.
amd64 as necessary. This is routine, except:
- the FAKE_MCOUNT($bintr) in doreti was missing the '$'. This gave a
a garbage address made up of padding bytes (with the nop byte 0x90 as
the MSB) instead of the intended address of bintr. This accidentally
worked on i386's because (0x90 << 24) is close enough to bintr, but
it doesn't work on amd64's because (0x90 << 56) is much further away
from bintr.
- the FAKE_MCOUNT($btrap) in calltrap was similarly broken. It hasn't
been needed since FreeBSD-1, so just delete it.
and high resolution profiling of interrupt handlers. The adjustments
are routine once the magic stack offset 13*4 is decoded to be TF_RIP
(there were originally more types of stack frames so using TF_EIP for
one of them wouldn't have been much simpler).
Removed garbage comments attached to some of the FAKE_MCOUNT()s.
that the usual macro for "ret" hides the detail of calling .mexitcount
before returning.
Fixed missing call to .mexitcount in lgdt(). This was missing on
i386's, mainly because lgdt() uses lret[q] insted of ret. This is
very unimportant since lgdt() is not (normally?) called until after
profiling is initialized.
in all USB ethernet drivers. The qdat structure contains a pointer
to the interface's struct ifnet and is used to process incoming
packets, so simultaneous use of two similar devices caused crashes
and confusion.
The if_udav driver appeared in the tree since Daan's PR, so I made
similar changes to that driver too.
PR: kern/59290
Submitted by: Daan Vreeken <Danovitsch@Vitsch.net>
MFNetBSD 1.177; author: toshii
Use the correct wValue to get hub desriptors.
Also, make wValue checks of root hub codes less strict.
MFNetBSD 1.178: author: martin
Interrupt descriptors might become invalid while being processed in
uhci_check_intr - so remember their next pointer before calling it.
Patch provided by Matthew Orgass in PR kern/24542.
Obtained from: NetBSD
(if trying to match only one real-group or one session-id), now that
those options are implemented in src/sys/kern/kern_proc.c (v1.203).
PR: bin/65803 (a very tiny piece of the PR)
Submitted by: Cyrille Lefevre
KERN_PROC_SESSION option which had been previously defined but
never implemented.
PR: bin/65803 (a very tiny piece of the PR)`
Submitted by: Cyrille Lefevre
Note that bsdtar's -o (which follows SUSv2) is not the same as GNU tar's -o.
In GNU tar, -o and --no-same-owner are not synonyms.
Pointed out by: Kris Kennaway (required by xpenguins port)
return value for BUS_READ_IVAR and thus don't generate the proper NULL
in cases where a device (i.e. on PCI) does not have a handle.
Found by: peadar, tjr
stored in EEPROM or NVRAM. It's inspired by the NetBSD eeprom(8) and
the SunOS/Solaris eeprom(1M) utilities. Currently, this eeprom(8)
only supports systems equipped with Open Firmware and is only tested
on Sun machines but should work on any platform using Open Firmware.
A bit more specific, eeprom(8) can be used on these systems to do the
same under FreeBSD as can be done using the printenv and setenv
commandos in the boot monitor. One thing that only hardly can be done
using the boot monitor but easily with eeprom(8) is to write a logo
to the "oem-logo" property. eeprom(8) may also be useful to recover
the boot monitor password (in the default configuration only as root,
of course), i.e. when the boot monitor allows you to boot but you
can't alter the configuration because the password is unknown. The
man page may also be a useful reference of the various configuration
variables.
The idea of eeprom(8) is that handlers can be written to add support
for any firmware that stores such configuration in EEPROM or NVRAM;
sort of e.g. eeprom(1M) on Solaris/x86 is used to turn PAE-support
on and off (stored in a file then, not hardware). In FreeBSD, a
candidate for this would be a handler for the EFI boot environment
for FreeBSD/ia64.
eeprom(8) uses some code from NetBSD (eeprom.c and the base for
eeprom.8), the handler for the Open Firmware /options node
(ofw_options.[c,h]) was written using ofw_util.[c,h] from ofwdump(8).
Reviewed by: ru (slightly earlier version of the man page)
- Make the code use the new OFIOCMAXVALUE instead of defining the maximum
length of property values locally.
- Move the application specific parts from ofw_util.c to ofwdump.c in
order to make ofw_util.c more library-like. While ofw_dump_properties()
could be made non-specific to ofwdump(8) it's currently optimized for
use in ofwdump(8) and making it a library-like function would just
complicate the code unnecessarily.
- Minor clean-up in ofw_util.c, e.g. make its use of getopt(3) the way
it's described in style(9), make its usage() static, etc.
- Add a comment in ofw_util.c about why it doesn't call usage() when
neither the "-a" option nor a node-name where given.
- Add ofw_optnode() and ofw_setprop(), helper functions for the
OFIOCGETOPTNODE and OFIOCSET ioctls respectively, to ofw_util.[c,h].
- Be consistent with the use of 'const' in ofw_util.[c,h] and add 'const'
to the function arguments that are acutally const but weren't declared
as such.
- Mark WARNS=6 clean.
Approved by: tmm
