Commit Graph

150863 Commits

Author SHA1 Message Date
Jung-uk Kim
35012a1e69 Add an experimental and rudimentary JIT optimizer to reduce unncessary
overhead from short BPF filter programs such as "get the first 96 bytes".
2009-11-21 00:19:09 +00:00
Konstantin Belousov
080136212f On the return path from F_RDAHEAD and F_READAHEAD fcntls, do not
unlock Giant twice.

While there, bring conditions in the do/while loops closer to style,
that also makes the lines fit into 80 columns.

Reported and tested by:	dougb
2009-11-20 22:22:53 +00:00
Rick Macklem
086f6e0cc7 Patch the experimental NFS server is a manner analagous to
r197525, so that the creation verifier is handled correctly
in va_atime for 64bit architectures. There were two problems.
One was that the code incorrectly assumed that
sizeof (struct timespec) == 8 and the other was that the tv_sec
field needs to be assigned from a signed 32bit integer, so that
sign extension occurs on 64bit architectures. This is required
for correct operation when exporting ZFS volumes.

Reviewed by:	pjd
MFC after:	2 weeks
2009-11-20 21:21:13 +00:00
Jung-uk Kim
c12b965f99 General style cleanup, no functional change. 2009-11-20 21:12:40 +00:00
John Baldwin
4e1fc19cea Revert the previous change to pthread_once() stub in libc. It is actually
a feature that libstdc++ depends on to simulate the behavior of libc's
internal '__isthreaded' variable.  One benefit of this is that _libc_once()
is now private to _once_stub.c.

Requested by:	kan
2009-11-20 20:43:34 +00:00
Pyun YongHyeon
ed848e3a02 Only Tx checksum offloading is supported now. Remove experimental
code sneaked in r199611.
2009-11-20 20:43:16 +00:00
Pyun YongHyeon
fe42b04d70 Add __FBSDID. 2009-11-20 20:40:34 +00:00
Pyun YongHyeon
9955274c64 Add IPv4/TCP/UDP Tx checksum offloading support. It seems the
controller also has support for IP/TCP checksum offloading for Rx
path. But I failed to find to way to enable Rx MAC to compute the
checksum of received frames.
2009-11-20 20:33:59 +00:00
Pyun YongHyeon
abd12fc6d6 Because we know received bytes including CRC there is no reason to
call m_adj(9). The controller also seems to have a capability to
strip CRC bytes but I failed to activate this feature except for
loopback traffic.
2009-11-20 20:25:21 +00:00
Pyun YongHyeon
26e07b50c6 Add initial endianness support. It seems the controller supports
both big-endian and little-endian format in descriptors for Rx path
but I couldn't find equivalent feature in Tx path. So just stick to
little-endian for now.
2009-11-20 20:18:53 +00:00
Pyun YongHyeon
bbd5260f13 Remove unnecessary structure packing. 2009-11-20 20:12:37 +00:00
John Baldwin
21b9a000c7 Replace gmt_is_set and the gmt_mutex lock with a pthread_once_t variable and
an init routine run on the first invocation via _once().

MFC after:	1 week
2009-11-20 19:21:33 +00:00
John Baldwin
5bcfe82e47 Add an internal _once() method. This works identical to pthread_once(3)
with the additional property that it is safe for routines in libc to use
in both single-threaded and multi-threaded processes.  Multi-threaded
processes use the pthread_once() implementation from the threading library
while single-threaded processes use a simplified "stub" version internal
to libc.  The libc stub-version of pthread_once() now also uses the
simplified "stub" version as well instead of being a nop.

Reviewed by:	deischen, Matthew Fleming @ Isilon
Suggested by:	alc
MFC after:	1 week
2009-11-20 19:19:51 +00:00
Jung-uk Kim
908d88b03b Correct copyright date in the BPF regression test. 2009-11-20 19:01:09 +00:00
Jung-uk Kim
448a7ff2ec Adjust BPF JIT compiler regression tests to catch up with r199603. 2009-11-20 18:53:38 +00:00
Jung-uk Kim
5ecf77367c - Allocate scratch memory on stack instead of pre-allocating it with
the filter as we do from bpf_filter()[1].
- Revert experimental use of contigmalloc(9)/contigfree(9).  It has no
performance benefit over malloc(9)/free(9)[2].

Requested by:	rwatson[1]
Pointed out by:	rwatson, jhb, alc[2]
2009-11-20 18:49:20 +00:00
Marcel Moolenaar
0c8e3b18f0 Always allocate PCI/ISA interrupts as shareable so that shared
interrupts don't cause driver attach failures.
2009-11-20 17:59:50 +00:00
Alexander Leidinger
9bf9af1004 Fix minor memory leak in a function.
MFC after:	1 week
2009-11-20 15:28:38 +00:00
Alexander Leidinger
4b2d15efe5 Fix minor resource leak in a function which was introduced by changing an
err() to a return in r106254.

MFC after:	1 week
2009-11-20 15:27:00 +00:00
Rafal Jaworowski
04311706d0 tsec: Use IFQ_DRV macros for managing interface packet queue.
This lets tsec(4) work with ALTQ.

Submitted by:	Marcin Ligenza
MFC after:	1 week
2009-11-20 13:28:06 +00:00
John Baldwin
c6858769af Always use 64-bit LBAs for disk addresses in zfsboot and gptzfsboot to
fully support booting from large volumes.

Tested by:	Emil Smolenski  ambsd of raisa.eu.org
Submitted by:	Matt Reimer  mattjreimer of gmail (most of the C bits)
MFC after:	1 week
2009-11-20 12:48:35 +00:00
Andrew Thompson
7fb4357098 remove volume alignment (was previously not correctly implemented)
Submitted by:	HPS
Reported by:	Jaakko Heinonen
2009-11-20 09:00:38 +00:00
Andrew Thompson
4594d907db Sync to P4
- fix a transfer cancelling bug/segfault [1]
 - correct a return code in the transfer cancel function.
 - add new API function, libusb20_tr_bulk_intr_sync().

Submitted by:	HPS
Reported by:	Robert Jenssen [1]
2009-11-20 08:57:25 +00:00
Marcel Moolenaar
1c8a163c8b No need to include opt_kstack_pages.h, because KSTACK_PAGES is
already defined through genassym.c
2009-11-20 07:40:02 +00:00
Marcel Moolenaar
02b5a86f38 Add a seatbelt to the Nested TLB Fault handler to give us a chance
to panic when we have an unexpected TLB fault while interrupt
collection is disabled. Use a token rather than the actual address
of the restart point to avoid the need for the movl instruction.
The token is arbitrary. For the drummers: it's based on a single
paradiddle.
2009-11-20 03:14:54 +00:00
Pyun YongHyeon
01d1a6c355 Move interface reinitialization down after disabling WOL in resume
path.
2009-11-19 23:14:40 +00:00
Pyun YongHyeon
8476c243bb Minimize interface reinitialization by checking IFF_DRV_RUNNING
flag. This fixes unnecessary interface UP/DOWNs during getting an
IP address via DHCP.

Tested by:	Warren Block ( wblock<> wonkity dot com )
2009-11-19 23:12:44 +00:00
Pyun YongHyeon
accb4fcd92 Fix copy & paste error and remove extra space before colon.
Pointed out by: danfe
2009-11-19 22:59:52 +00:00
Pyun YongHyeon
8b3c64969b Use capability pointer to access PCIe registers rather than
directly access them at fixed address. Frequently the register
offset could be changed if additional PCI capabilities are added to
controller.
One odd thing is ET_PCIR_L0S_L1_LATENCY register. I think it's PCIe
link capabilities register but the location of the register does
not match with PCIe capability pointer + offset. I'm not sure it's
shadow register of PCIe link capabilities register.
2009-11-19 22:53:41 +00:00
John Baldwin
e55bc0154d - Hook into the existing stat timer to drive the transmit watchdog instead
of using if_watchdog and if_timer.
- Reorder detach to call ether_ifdetach() before anything else in tl(4)
  and wb(4).
2009-11-19 22:14:23 +00:00
John Baldwin
7cf545d0a1 - Add a private timer to drive the transmit watchdog instead of using
if_watchdog and if_timer.
- Fix some issues in detach for sn(4), ste(4), and ti(4).  Primarily this
  means calling ether_ifdetach() before anything else.
2009-11-19 22:06:40 +00:00
Pyun YongHyeon
d95121b427 Use bus_{read,write}_4 rather than bus_space_{read,write}_4. 2009-11-19 22:06:19 +00:00
John Baldwin
3064f0530b - Initialize callout before it is used in atestop() during attach.
- Reorder detach so that ether_ifdetach() is called first.  This removes
  the race that ATE_FLAG_DETACHING closed, so that flag can be removed.
- Trim a duplicate clearing of IFF_DRV_RUNNING.

Reviewed by:	imp
2009-11-19 22:04:02 +00:00
Pyun YongHyeon
398f1b6501 style(9) 2009-11-19 21:53:21 +00:00
John Baldwin
d31c51daa7 Fix compile after previous "harmless" commit.
Pointy hat to:	jhb
2009-11-19 21:47:54 +00:00
Pyun YongHyeon
73e338b028 Remove extra spce at the EOL. 2009-11-19 21:46:58 +00:00
Pyun YongHyeon
cc3c3b4ec4 Add MSI support. 2009-11-19 21:45:06 +00:00
Pyun YongHyeon
5b8f4900f3 Destroy driver mutex in device detach. 2009-11-19 21:39:43 +00:00
Pyun YongHyeon
696e249300 Remove support code for FreeBSD 6.x versions. 2009-11-19 21:08:33 +00:00
John Baldwin
dbd69bc565 Remove commented out reference to if_watchdog and an assignment of zero to
if_timer.

Reviewed by:	scottl
2009-11-19 20:59:40 +00:00
Pyun YongHyeon
23263665d5 Remove complex macros that were used to compute bits values.
Although these macros may have its own strength, its complex
definition make hard to read the code.

Approved by:	delphij
2009-11-19 20:57:35 +00:00
John Baldwin
67e1dfa795 Several fixes to this driver:
- Overhaul the locking to avoid recursion and add missing locking in a few
  places.
- Don't schedule a task to call vge_start() from contexts that are safe to
  call vge_start() directly.  Just invoke the routine directly instead
  (this is what all of the other NIC drivers I am familiar with do).  Note
  that vge(4) does not use an interrupt filter handler which is the primary
  reason some other drivers use tasks.
- Add a new private timer to drive the watchdog timer instead of using
  if_watchdog and if_timer.
- Fixup detach by calling ether_ifdetach() before stopping the interface.
2009-11-19 19:35:15 +00:00
John Baldwin
71098608fc Several fixes to these drivers. Note that these two drivers are actually
just two different attachments (EISA and PCI) to a single driver.
- Add real locking.  Previously these drivers only acquired their lock
  in their interrupt handler or in the ioctl routine (but too broadly in
  the latter).  No locking was used for the stack calling down into the
  driver via if_init() or if_start(), for device shutdown or detach.  Also,
  the interrupt handler held the driver lock while calling if_input().  All
  this stuff should be fixed in the locking changes.
- Really fix these drivers to handle if_alloc().  The front-end attachments
  were using if_initname() before the ifnet was allocated.  Fix this by
  moving some of the duplicated logic from each driver into pdq_ifattach().
  While here, make pdq_ifattach() return an error so that the driver just
  fails to attach if if_alloc() fails rather than panic'ing.  Also, defer
  freeing the ifnet until the driver has stopped using it during detach.
- Add a new private timer to drive the watchdog timer.
- Pass the softc pointer to the interrupt handlers instead of the device_t
  so we can avoid the use of device_get_softc() and to better match what
  other drivers do.
2009-11-19 19:25:47 +00:00
John Baldwin
d4e576c497 - This driver used the if_watchdog timer both as a watchdog on transmit and
auto-negotiation.  To make this simpler and easier to understand I have
  split this out into two separate timers.  One just manages the auto-neg
  side of things and one is a transmit watchdog.  Neither uses if_watchdog.
- Call ether_ifdetach() at the start of detach.
2009-11-19 18:43:43 +00:00
John Baldwin
770c8c790f - Use callout_init_mtx() instead of callout_init(..., CALLOUT_MPSAFE).
- Add a missing callout_drain() to detach.
- Hook into the stats timer and use that to drive the transmit watchdog
  instead of using if_watchdog.
- Run the stats timer every second to match other drivers instead of every
  other second.
- Remove dubious callout handling that stopped the timer only to start it
  again while holding the driver lock without dropping it in between the
  stop and the start.
2009-11-19 18:37:55 +00:00
John Baldwin
73e7361eac This driver has two modes, a netgraph mode and an ifnet mode. In the
netgraph mode it used a private timer to drive the transmit watchdog.  In
the ifnet mode it used if_watchdog.  Now it always uses the private timer.
2009-11-19 18:21:51 +00:00
John Baldwin
e21e2eea58 These drivers only set if_timer but never set if_watchdog. Just remove
the assignments to if_timer.
2009-11-19 18:11:23 +00:00
Michael Reifenberger
569b2f6954 Add nanobsd example for the PC Engines ALIX board (serial console). 2009-11-19 16:27:51 +00:00
Alexander Motin
9e30fbf662 Tune CAM ATA kernel options a bit. Move PMP support from da to scbus and
add ada device option, according to man page.
2009-11-19 16:26:07 +00:00
Rafal Jaworowski
035622ef18 Provide an effective (relocated) address when building modules metadata.
This lets modules loaded dynamically in loader(8) work for U-Boot-based
platforms.

MFC after:	1 week
2009-11-19 16:25:41 +00:00