1
0
mirror of https://git.FreeBSD.org/src.git synced 2024-12-18 10:35:55 +00:00
Commit Graph

136107 Commits

Author SHA1 Message Date
Bruce Evans
9ed67737f2 Don't try to build s_nanl.c before it is committed. 2007-12-17 13:20:38 +00:00
Oleg Bulyzhin
d60810f26e Calculate p.fs.lookup_step correctly. This should prevent zeroing of
w_q_lookup table (used in RED algorithm for (1 - w_q)^t computation).

MFC after:	1 months
2007-12-17 10:25:56 +00:00
Kip Macy
c662126d0c disable update in place on transmit 2007-12-17 10:02:29 +00:00
Kip Macy
5e0f5cfaed Add SB_NOCOALESCE flag to disable socket buffer update in place 2007-12-17 10:02:01 +00:00
Diomidis Spinellis
3db8241bb4 Calling any function from vfork other than exec* and _exit yields
undefined behavior.

Noted by:	alfred
2007-12-17 09:02:42 +00:00
Kip Macy
8090c9f504 Make TCP offload work on HEAD (modulo negative interaction between sbcompress
and t3_push_frames).
 - Import latest changes to cxgb_main.c and cxgb_sge.c from toestack p4 branch
 - make driver local copy of tcp_subr.c and tcp_usrreq.c and override tcp_usrreqs so
   TOE can also functions on versions with unmodified TCP

- add cxgb back to the build
2007-12-17 08:17:51 +00:00
David Xu
7fab871d8c Check NULL pointer. 2007-12-17 08:09:37 +00:00
David Schultz
baa15be0ea Fix an int overflow on very large file systems.
PR:		bin/113399
Submitted by:	Staffan Ulfberg <staffan@ulfberg.se>
2007-12-17 08:03:18 +00:00
Kip Macy
228d1e266e build tcp_offload.c instead of tcp_ofld.c 2007-12-17 08:00:08 +00:00
Kip Macy
8b5709dfab incorporate feedback since initial commit
- rename tcp_ofld.[ch] to tcp_offload.[ch]
- document usage and locking conventions of the functions in the
  toe_usrreqs function vector
- document tcpcb, inpcb, and socket fields used by toe
- widen the listen interface into 2 functions
- rename DISABLE_TCP_OFFLOAD to TCP_OFFLOAD_DISABLE
- shrink conditional compilation to reduce the likelihood of bitrot
- replace sc->sc_toepcb checks in tcp_syncache.c with TOEPCB_ISSET
2007-12-17 07:56:27 +00:00
Kip Macy
29910a5a77 widen the routing event interface (arp update, redirect, and eventually pmtu change)
into separate functions

revert previous commit's changes to arpresolve and add a new interface
arpresolve2 which does arp resolution without an mbuf
2007-12-17 07:40:34 +00:00
Maxim Konovalov
fe832db659 o Fix a holiday to reflect a reality.
Submitted by:	Andrey V. Elsukov
MFC after:	1 week
2007-12-17 07:36:39 +00:00
David Xu
9514dcc041 Add missing changes for fixing LOR of umtx lock and thread lock, follow
the committing of files:
	kern_resource.c revision 1.181
	sched_4bsd.c	revision 1.111
	sched_ule.c	revision 1.218
2007-12-17 05:55:07 +00:00
Kevin Lo
5497f4c53e Use M_NOWAIT instead of M_WAITOK to cause malloc() to return NULL
Reviewed by: imp
2007-12-17 05:08:54 +00:00
Kip Macy
58505389d1 Don't panic in arpresolve if we're given a null mbuf. We could
insist that the caller just pass in an initialized mbuf even
if didn't have any data - but that seems rather contrived.
2007-12-17 04:19:25 +00:00
David Schultz
6821aba9e5 Add logbl(3) to libm. 2007-12-17 03:53:38 +00:00
David Xu
c5f411515c Add function prototypes. 2007-12-17 02:53:11 +00:00
Jason Evans
d55bd6236f Refactor features a bit in order to make it possible to disable lazy
deallocation and dynamic load balancing via the MALLOC_LAZY_FREE and
MALLOC_BALANCE knobs.  This is a non-functional change, since these
features are still enabled when possible.

Clean up a few things that more pedantic compiler settings would cause
complaints over.
2007-12-17 01:20:04 +00:00
David Schultz
3be0479b4c Document the fact that we have nan(3) now, and make some minor clarifications
in other places.
2007-12-17 01:04:43 +00:00
David Schultz
9cb4166881 On platforms where long is 64 bits, make sure gdtoa doesn't use a
long when it really wants an int.
2007-12-17 00:19:49 +00:00
David Schultz
5a97a86c28 Fix an amusing typo that has prevented this from compiling since 2004. 2007-12-16 23:38:55 +00:00
David Schultz
3f7112a1f4 Don't try the long double tests on i386. Our reduced precision
can cause them to fail.
2007-12-16 23:36:37 +00:00
David Schultz
fb048cca0c Fix previous commit: We should only mask the top bit in the STRTOG_NaNBits
case, not the STRTOG_Normal case.
2007-12-16 23:31:55 +00:00
Daniel Eischen
3d6d3ed091 Remove hacks to allow libkse to export its symbols in the LIBTHREAD_1_0
version namespace which was needed before the library version was
bumped.
2007-12-16 23:29:57 +00:00
David Schultz
64577b835b Remove another Alpha remnant. 2007-12-16 23:00:18 +00:00
Kip Macy
ea6ee7d60a Include cdefs.h and param.h for architectures with less header pollution 2007-12-16 21:22:24 +00:00
David Schultz
83f63fd648 Regression tests for nan{,f,l}(). 2007-12-16 21:19:51 +00:00
David Schultz
4b6b574455 Implement and document nan(), nanf(), and nanl(). This commit
adds two new directories in msun: ld80 and ld128. These are for
long double functions specific to the 80-bit long double format
used on x86-derived architectures, and the 128-bit format used on
sparc64, respectively.
2007-12-16 21:19:28 +00:00
Poul-Henning Kamp
e34f6b7311 Recognize CFA devices using either identification method. 2007-12-16 21:19:07 +00:00
Poul-Henning Kamp
164a5044a9 Add the magic word 0 value for CFA compliant devices. 2007-12-16 21:18:13 +00:00
David Schultz
39e7abef0e Export gdtoa's __ULto{x,Q}_D2A routine in a private namespace so
libm can use it.
2007-12-16 21:15:57 +00:00
David Schultz
199cdab56f Arrange so that the NaN returned by strtod("nan", NULL) is the same as
the NaN returned by strtod("nan()", NULL).
2007-12-16 21:15:09 +00:00
David Schultz
10a381e55a Fix strto{f,d,ld}() so they don't return signaling NaNs. C99
says they are never supposed to, and the fact that they did could
cause apps that run with unmasked FP exceptions to SIGFPE after a
scanf() or strtod(). The vendor stated that he will not be fixing
this, citing portability concerns.
2007-12-16 21:14:33 +00:00
David Schultz
2ca7a12a81 Some changes the vendor didn't want:
- Accept the '0x' prefix so strtod("nan(0x...)", NULL) returns the same
  thing as gcc's builtin nan("0x...") for such strings.
- Don't return uninitialized memory.
- Finish processing the string up to the closing ')' (provided it's
  lexically valid) for compatibility with C99 and *scanf().
2007-12-16 21:13:54 +00:00
Rui Paulo
de272a0a53 Fix previous commit. The code ended up in the wrong function.
Approved by:	     njl (mentor)
2007-12-16 20:37:27 +00:00
Poul-Henning Kamp
59c0f72857 Report erase interval (correctly) in sectors. 2007-12-16 20:19:55 +00:00
Poul-Henning Kamp
9a6378d803 Rename the undocumented -E option to -X.
Implement -E option which will erase the filesystem sectors before
making the new filesystem.  Reserved space in front of the superblock
(bootcode) is not erased.

NB: Erasing can take as long time as writing every sector sequentially.

This is relevant for all flash based disks which use wearlevelling.
2007-12-16 19:41:31 +00:00
Poul-Henning Kamp
015a11e695 Chop DIOCGDELETE from userland up in 1024 sector chunks to give geom_disk
or any other bio chopping geom a reasonable size of work.

Check for delivered signals between chunks, because the request size
and service time is unbounded.
2007-12-16 19:38:26 +00:00
Kip Macy
bfa0795b00 Add back in cxgb now that the includes are believed to be fixed 2007-12-16 18:45:23 +00:00
Kip Macy
fc5a2e51fb Use the vm include convention of busdma 2007-12-16 18:43:57 +00:00
Kip Macy
04b25f8e3e need M_IOVEC define 2007-12-16 18:36:57 +00:00
Kip Macy
01cf8d43b8 Don't globally include mvec.h its only needed by cxgb_sge.c 2007-12-16 18:26:04 +00:00
Poul-Henning Kamp
eed6cda966 Don't limit BIO_DELETE requests to MAXPHYS, they perform no data
transfers, so they are not subject to the VM system limitation.
2007-12-16 18:03:31 +00:00
Poul-Henning Kamp
20a0f65b77 Add a berase() function which uses ioctl(DIOCGDELETE) to erase a slab
of the disk.
2007-12-16 18:02:37 +00:00
Diomidis Spinellis
2ec43f35f4 Eliminate gcc "variable clobbered" warnings by declaring the variables
living across the vfork as volatile.

Noted by:	kan
2007-12-16 17:47:34 +00:00
Kip Macy
3bd3bd76b5 Remove cxgb module from build until I can compensate for the lack of header
pollution on architectures other than amd64
2007-12-16 17:43:40 +00:00
Poul-Henning Kamp
9d599522c3 Report CFA extension, CFA -> (Compact) Flash Association 2007-12-16 15:11:40 +00:00
Diomidis Spinellis
3a44a28835 When moving a directory across devices to a place where a directory
with the same name exists, delete that directory first, before performing
the copy.  This ensures that mv(1) across devices follows the semantics
of rename(2), as required by POSIX.

This change could introduce the potential of data loss, even if the
copy fails, violating the atomicity properties of rename(2).  This is
(mostly) mitigated by first renaming the destination and obliterating
it only after a succesfull copy.

The above logic also led to the introduction of code that will cleanup
the results of a partial copy, if a cross-device copy fails.

PR:		bin/118367
MFC after:	1 month
2007-12-16 14:14:31 +00:00
Bruce A. Mah
720788ca65 Autogenerate hardware notes for hptrr(4). 2007-12-16 13:14:47 +00:00
Stanislav Sedov
1e6774a44e - Don't return 0xffff if PHY id isn't equal 0. This allows PHYs with
non-zero addresses to be used.

Approved by:	cognet
MFC after:	2 weeks
2007-12-16 12:57:12 +00:00