Peter Wemm [Fri, 6 Jul 2007 00:47:44 +0000 (00:47 +0000)]
Fix warnings.
nxge: cast page size fragments down to (int). If the vm's demand paging
PAGE_SIZE is ever too big for that, we've got far bigger problems.
ofw: move va_start() a little earlier. gcc-4.2 doesn't like us modifying
the last arg before the va_start().
re(4) devices requires an external EEPROM. Depending on models it
would be 93C46(1Kbit) or 93C56(2Kbit). One of differences between them
is number of address lines required to access the EEPROM. For example,
93C56 EEPROM needs 8 address lines to read/write data. If 93C56
recevied premature end of required number of serial clock(CLK) to set
OP code/address of EEPROM, the result would be unexpected behavior.
Previously it tried to detect 93C46, which requires 6 address lines,
and then assumed it would be 93C56 if read data was not expected
value. However, this approach didn't work in some models/situations
as 93C56 requries 8 address lines to access its data. In order to fix
it, change EEPROM probing order such that 93C56 is detected reliably.
While I'm here change hard-coded address line numbers with defined
constant to enhance readability.
Robert Watson [Thu, 5 Jul 2007 13:16:04 +0000 (13:16 +0000)]
In preparation for 7.0 privilege cleanup, clean up style:
- Sort copyrights by date.
- Re-wrap, and in some cases, fix comments.
- Fix tabbing, white space, remove extra blank lines.
- Remove commented out debugging printfs.
Andrew Thompson [Thu, 5 Jul 2007 09:18:57 +0000 (09:18 +0000)]
Allow the LACP state to be queried from userland which at the moment is the
actor and partner peer info. Print out the active aggregator and per port data
in verbose mode from ifconfig.
This was needed during the IPSEC->FAST_IPSEC->IPSEC transition
period to not break the build after picking up netipsec header
files. Now that the FAST_IPSEC kernel option is gone and the
default is IPSEC again those defines are superfluous.
Remove netkey directory from cscope/TAGs generation and replace
it with netipsec now that KAME IPsec is gone.
While here add missing netinet6 directories.
Add comments about the ports needed to be able to run those targets.
Peter Wemm [Thu, 5 Jul 2007 07:46:33 +0000 (07:46 +0000)]
Quiet framelen uninitialized warning. I think it was a false alarm.
If check_fhdr() returns false, the frame_ok variable should protect any
meaningful evaluations of framelen.
Peter Wemm [Thu, 5 Jul 2007 06:59:14 +0000 (06:59 +0000)]
Quiet warnings. These do not appear to be actually used uninitialized,
but gcc's optimizer isn't smart enough to see that. Pre-initializing
seems harmless enough.
Warner Losh [Thu, 5 Jul 2007 05:26:08 +0000 (05:26 +0000)]
Add quirks, derived from the linux quirks table for:
o Adonics Cable 205
o Aiptek PocketCAM 3Mega
o Belkin USB2SCSI
o Casio QV DigiCam
o CCYU EasyDisk ED1064
o Desknote UCR-61S2B
o Epson Stylus Photo 875DC Card Reader
o Epson Stylus Photo 895 Card Reader
o Feiya 5-in-1 Card Reader
o Hitachi Dvd-CAM DZ-MV100A Camcorder
o HP CD-WRiter+ CD-4e
o Insystem Storage Adapter v2
o Kyocera Finecam S3x
o Kyocera Finecam S4
o Kyocera Finecam S5
o Kyocera Finecam L3
o Lexar USB CF Reader
o MindAtWork Digital Wallet
o Minolta Dimage F300
o Minolta Dimage E223
o Minsumi USB Fdd
o Netac USB-CF-Card
o NetChip USB Clik! 40
o Onspec MDCFE-B USB CF Reader
o Onspec SIIG/Datafab Memory Stick + CF Reader/Writer
o Onspec Datafab-based Reader
o Onspec PNY/Datafab CF+SM Reader
o Onspec SimpleTech/Datafab CF+SM Reader
o Onspec MDSM-b Reader
o Onspec USB To CF + SM Combo (LC1)
o Onspec ImageMate SDDR55
o Panasonic LS-120 Camera
o Samsung Techwin Digimax 410
o Shuttle eUSB SmartMedia / CompactFlash Adapter
o Skanhex MD 7425 Camera
o Skanhex SX 520z Camera
o Sony Memorystick NW-MS7
o Sony Portable USB Hardrive V2
o Sony Memorystick PEG N760c
o Sony Memorystick MSC-U03
o TREK/IBM USB memory key
o Trumpion T33520 USB Flash Card Controller
o Trumpion MP3 Player
o Vivtar Vivicam 35Xx
o WinMaxGroup USB Flash Disk 64M-C
o Zoran Digital Camera EX-20 DSC
and maybe a few others...
Submitted by: Vaidas Damosevicius and flz
PR: 79893
Reviewed by: njl, flz
Approved by: re (blanket)
Peter Wemm [Wed, 4 Jul 2007 23:27:38 +0000 (23:27 +0000)]
Change the C wrappers for mmap/lseek/pread/pwrite/truncate/ftruncate to
call the pad-less versions of the corresponding syscalls if the running
kernel supports it. Check kern.osreldate once per program and cache the
result to select the appropriate syscall. This maintains userland
compatability with kernel.old's from quite a while back.
Peter Wemm [Wed, 4 Jul 2007 23:23:01 +0000 (23:23 +0000)]
Classify mmap, lseek, pread, pwrite, truncate, ftruncate as pseudo
syscalls, unless WITHOUT_SYSCALL_COMPAT is defined. The default case
will have the .c wrappers still. If you define WITHOUT_SYSCALL_COMPAT,
the .c wrappers will go away and libc will make direct syscalls.
After 7-stable starts, the direct syscall method will be default.
Peter Wemm [Wed, 4 Jul 2007 23:18:38 +0000 (23:18 +0000)]
Adjust the syscall stub macros to be consistent in their meaning. In
particular:
SYSCALL() makes a syscall, with errno handling, and continues execution
directly after the macro in the non-error case.
RSYSCALL() is just like SYSCALL(), but returns after success.
Both SYSCALL(name) and RSYSCALL(name) export "__sys_name" as a strong
symbol, with "_name" and "name" as weak aliases.
PSEUDO() is just like RSYSCALL(), but skipping the "name" weak alias. It
still does "__sys_name" and "_name".
Change i386 to add errno handling to PSEUDO. The same for amd64 and
sparc64, with appear to have copied the behavior.
ia64 was correct (as was alpha). Just remove some apparently unused
variants of the macros. (untested!)
I believe powerpc is correct.
Fix arm to not export "name" from the PSEUDO case. Remove apparently
extra unused variants. (untested!)
The errno problem manifested on i386/amd64/sparc64 by having "PSEUDO"
classified syscalls return without setting errno. eg: "addr = mmap()"
could return with "addr" = 22 instead of setting errno to 22 and
returning -1.
Peter Wemm [Wed, 4 Jul 2007 23:02:40 +0000 (23:02 +0000)]
Add i386 emulation wrappers for mmap/lseek/etc. These use COMPAT6, so
you must use the already existing, already in generic, COMPAT_FREEBSD6
kernel option for running old 32 bit binaries.
Peter Wemm [Wed, 4 Jul 2007 22:47:37 +0000 (22:47 +0000)]
Create new syscalls for mmap(), lseek(), pread(), pwrite(), truncate() and
ftruncate(), but without the pad arg.
There are several reasons for this. Consider 'mmap()'. On AMD64, the
function call (and syscall) ABI allow for 6 register arguments. Additional
arguments go on the stack. mmap(2) has 6 arguments. However, the syscall
definition has an extra 'int pad' argument. This pushes it to 7 arguments,
which means one must spill into the memory stack. Since the kernel API
doesn't match userland API, we have a hack in libc - libc/sys/mmap.c.
This implements the userland API by calling __syscall() with an extra
argument and the pad argument, for a total of 8 args. This is all
unnecessary and inconvenient for several things, including the kernel's
syscall handler code which now has to handle merging stack arguments with
register arguments. It is a big deal for certain 3rd party code.
I'm adding libc glue to make the transition totally painless. I had
intended to mark the old syscalls as COMPAT6, but the potential to shoot
your feet by building a new kernel without COMPAT_FREEBSD6 but with a
slighly older userland was too great. For now, they have manual
"freebsd6_" prefixes rather than being COMPAT6. They will go back to
being marked 'COMPAT6' after 7-stable starts.
Peter Wemm [Wed, 4 Jul 2007 22:38:28 +0000 (22:38 +0000)]
Add support for COMPAT6 syscalls.
Also, change the visibility of compat syscalls a slightly. Compat
syscalls were missing from 'syscalls.h' entirely. This additionally adds
them with their compat prefix. eg: SYS_freebsd6_mmap.
Also, the syscalls.c names strings have different prefixes to differentiate
syscalls. Instead of several "old.mmap" strings, there will now be a
"compat.mmap" and "compat6.mmap" etc. Before, both would have had the
same "old.mmap" label.
Ariff Abdullah [Wed, 4 Jul 2007 12:33:11 +0000 (12:33 +0000)]
Be much more forgiving towards applications that requesting ioctls
that should be a no-op (for example, requesting SYNC on record path).
The standards does not indicate that such requests are illegal, so
just return it as success instead of EINVAL.
Since cdev mutex is after system map mutex in global lock order, free()
shall not be called while holding cdev mutex. devfs_inos unrhdr has cdev as
mutex, thus creating this LOR situation.
Postpone calling free() in kern/subr_unit.c:alloc_unr() and nested functions
until the unrhdr mutex is dropped. Save the freed items on the ppfree list
instead, and provide the clean_unrhdrl() and clean_unrhdr() functions to
clean the list.
Call clean_unrhdrl() after devfs_create() calls immediately before
dropping cdev mutex. devfs_create() is the only user of the alloc_unrl()
in the tree.
Reviewed by: phk
Tested by: Peter Holm
LOR: 80
Approved by: re (kensmith)
Peter Wemm [Wed, 4 Jul 2007 00:55:50 +0000 (00:55 +0000)]
Fix missing prototype warnings. (Compile errors with -Werror on)
When using namespace.h/un-namespace.h, you use _ versions of syscalls.
Change getsockopt() to _getsockopt() and same for setsockopt().
Sean Farley [Wed, 4 Jul 2007 00:00:41 +0000 (00:00 +0000)]
Significantly reduce the memory leak as noted in BUGS section for
setenv(3) by tracking the size of the memory allocated instead of using
strlen() on the current value.
Convert all calls to POSIX from historic BSD API:
- unsetenv returns an int.
- putenv takes a char * instead of const char *.
- putenv no longer makes a copy of the input string.
- errno is set appropriately for POSIX. Exceptions involve bad environ
variable and internal initialization code. These both set errno to
EFAULT.
Several patches to base utilities to handle the POSIX changes from
Andrey Chernov's previous commit. A few I re-wrote to use setenv()
instead of putenv().
New regression module for tools/regression/environ to test these
functions. It also can be used to test the performance.
Bump __FreeBSD_version to 700050 due to API change.
PR: kern/99826
Approved by: wes
Approved by: re (kensmith)
Jeff Roberson [Tue, 3 Jul 2007 21:26:06 +0000 (21:26 +0000)]
- Use explicit locking in the various fcntl case statements so that we
can acquire shared filedescriptor locks in the appropriate cases.
- Remove Giant from calls that issue ioctls. The ioctl path has been
mpsafe for some time now.
- Only acquire giant for VOP_ADVLOCK when the filesystem requires giant.
advlock is now mpsafe.
Jeff Roberson [Tue, 3 Jul 2007 21:22:58 +0000 (21:22 +0000)]
- Remove explicit Giant protection from lockf. Use the vnode interlock
to protect this datastructure instead.
- Preallocate an extra lockf structure in case we want to split a lock
on insert or delete.
- msleep() on the vnode interlock when blocking on a lock.
John Baldwin [Tue, 3 Jul 2007 18:37:06 +0000 (18:37 +0000)]
Tweak the low-level MI SMP code some:
- Use cpu_spinwait() in the spin loops in stop_cpus(), restart_cpus(), and
smp_rendezvous_action().
- Remove unneeded acq memory barriers in stop_cpus(), restart_cpus(), and
smp_rendezvous_action().
- Add an additional synch point in smp_rendezvous() to ensure that all the
CPUs will always see an up-to-date value of smp_rv_setup_func.
Reviewed by: attilio
Approved by: re (kensmith)
Tested on: alpha, amd64, i386, sparc64 SMP (for several years)
John Baldwin [Tue, 3 Jul 2007 18:31:47 +0000 (18:31 +0000)]
Fix for a race where out of order loading of NFS attrs into the
nfsnode could lead to attrs being stale. One example (that we
ran into) was a READDIR+, WRITE. The responses came back in
order, but the attrs from the WRITE were loaded before the
attrs from the READDIR+, leading to the wrong size from being
read on the next stat() call.
John Baldwin [Tue, 3 Jul 2007 18:30:55 +0000 (18:30 +0000)]
Fix up NFS client write error handling. Errors are split into
recoverable and unrecoverable. For the former, we redirty the
buffer and hang onto it for future retries. For the latter (eg.
ESTALE), we discard the buffer and return the error back to the
user on the next syscall. This fixes a number of vfs panics and
fixes having a large number of dirty buffers (that cannot be
written out and reclaimed) from hanging around. Thanks to ups@
for discussions on this issue.
Reported by: kris, Kai, others
Approved by: re (kensmith)
Rev. 1.204 and 1.205 got an erronous version of destroy_dev() that
calls destroy_dev_sched() with cdev mutex locked. Commit the code
that was actually tested.
Lock Giant and proctree lock around dereferencing p_session->s_ttyvp->v_rdev.
Lock cdev mutex too to close the race with tty being freed.
Relock clone_drain_lock to prevent the LOR with proctree lock, thus
add #include <fs/devfs/devfs_int.h>.
Suggested by: tegge
Debugging help and testing by: Peter Holm
Approved by: re (kensmith)
Use make_dev_credf(MAKEDEV_REF) instead of make_dev() from the clone handler.
Lock Giant in the clone handler.
Use destroy_dev_sched() explicitely from pty_maybecleanup() and postpone
pty_release() until both master and slave cdevs are destroyed by setting
it as callback for destroy_dev_sched().
Debugging help and testing by: Peter Holm
Approved by: re (kensmith)
Automatically detect deadlock condition in destroy_dev(), that is, if
destroy_dev() is called from csw method, and no d_purge driver method is
provided. Transform the direct call to destroy_dev() into destroy_dev_sched().
Reviewed by: njl (programming interface)
Debugging help and testing by: Peter Holm
Approved by: re (kensmith)
Since rev. 1.199 of sys/kern/kern_conf.c, the thread that calls
destroy_dev() from d_close() cdev method would self-deadlock.
devfs_close() bump device thread reference counter, and destroy_dev()
sleeps, waiting for si_threadcount to reach zero for cdev without
d_purge method.
destroy_dev_sched() could be used instead from d_close(), to
schedule execution of destroy_dev() in another context. The
destroy_dev_sched_drain() function can be used to drain the scheduled
calls to destroy_dev_sched(). Similarly, drain_dev_clone_events() drains
the events clone to make sure no lingering devices are left after
dev_clone event handler deregistered.
make_dev_credf(MAKEDEV_REF) function should be used from dev_clone
event handlers instead of make_dev()/make_dev_cred() to ensure that created
device has reference counter bumped before cdev mutex is dropped inside
make_dev().
Reviewed by: tegge (early versions), njl (programming interface)
Debugging help and testing by: Peter Holm
Approved by: re (kensmith)
Matt Jacob [Mon, 2 Jul 2007 20:08:20 +0000 (20:08 +0000)]
Recover from some major omissions/problems with the 24XX port.
First, we were never correctly checking for a 24XX Status Type 0
response- that cased us to fall through to evaluate status for
commands as if this were a 2100/2200/2300 Status Type 0 response.
This is *close*, but not quite the same. This has been reported
to be apparent with some wierd lun configuration problems with
some arrays. It became glaringly apparent on sparc64 where none
of the correct byte swap things were done.
Fixing this omission then caused a whole universe shifting debug
cycle of endian issues for the 2400. The manual for 24XX f/w turns
out to be wrong about the endianness of a couple of entities. The
lun and cdb fields for the type 7 request are *not* unconditionally
big endian- they happen to be opposite of whatever the endian of
the current machine type is. Same with the sense data for the
24XX type 0 response.
While we're at it investigate and resolve some NVRAM endian
issues.
Randall Stewart [Mon, 2 Jul 2007 19:22:22 +0000 (19:22 +0000)]
- Consolidate the code that free's chunks to actually also
call the sctp_free_remote_address() function.
- Assure that when we allocate a chunk the whoTo is NULL,
also when we free it and place it into the cache we NULL
it (that way the consolidation code will always work).
- Fix a small race, when a empty data holder is left on the stream
out queue, and both sides do a shutdown, the empty data holder
would prevent us from sending a SHUTDOWN-ACK and at the same time we
never would cleanup the empty holder (since nothing was ever in queue).
We now add a utility function that a) cleans up empty holders and
b) properly determines if there are still pending data chunks on
the stream out wheel.
Approved by: re@freebsd.org (Ken Smith)
Robert Watson [Mon, 2 Jul 2007 14:03:29 +0000 (14:03 +0000)]
Continue kernel privilege cleanup for 7.0: unstaticize suser_enabled and
stop declaring it in systm.h -- it's used only in kern_priv.c and is not
required elsewhere.