ı`151652 183 361 195 729 114 178 133 3081 197 195 265 129 166 189 189 232 232 198 201 166 169 191 196 152 156 132 134 735 135 178 189 353 221 133 181 178 370 288 134 134 1172 607 221 195 274 170 182 1020 316 237 232 190 1408 143 263 739 283 341 239 348 188 180 187 298 377 316 187 155 210 286 151 349 427 336 450 188 287 160 142 228 380 245 170 234 241 294 312 314 183 932 265 219 166 656 110 226 389 374 195 312 234 185 197 232 313 229 180 165 242 189 169 173 140 433 436 204 184 250 519 551 237 725 362 611 343 393 315 420 584 702 188 147 158 122 165 115 162 120 139 224 351 170 342 316 593 1081 299 161 185 136 179 164 166 165 438 383 210 159 221 297 304 384 144 289 123 144 159 169 169 170 165 321 346 206 168 216 159 422 333 432 283 334 221 229 K 10 svn:author V 5 wpaul K 8 svn:date V 27 2005-10-25T16:14:34.000000Z K 7 svn:log V 266 Correct some __FreeBSD_version conditionals to use version bumps closer to the actual dates when code actually changed. Also add special case link state change handling for RELENG_5, which doesn't have if_link_state_change(). No actual operational changes are done. END K 10 svn:author V 5 wpaul K 8 svn:date V 27 2005-10-25T16:16:01.000000Z K 7 svn:log V 100 MFC: correct some __FreeBSD_version conditionals, fix link state change notifications for RELENG_5. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-10-25T18:47:12.000000Z K 7 svn:log V 632 Updates and enhancements to the multi-threaded httpd performance test tool: - Use uname(3) to query the OS name to report in the HTTP headers. This is probably more useful than hard-coding FreeBSD. - If no path is specified, create a 1k temporary file and send that instead. Pass a file descriptor into http_serve() rather than using a global fd. - Add more carriage returns to the HTTP headers to be a bit more correct. (Suggested by: andre) - Read to a buffer rather than a single character to reduce the number of recv() system calls pulling in the HTTP request. - Properly wait for two, not one, \n's on input. END K 10 svn:author V 8 brueffer K 8 svn:date V 27 2005-10-25T18:57:03.000000Z K 7 svn:log V 17 Xref polling(4). END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-25T19:29:02.000000Z K 7 svn:log V 86 Use shorter names for the Giant and fast taskqueues so that their names actually fit. END K 10 svn:author V 7 delphij K 8 svn:date V 27 2005-10-25T19:46:15.000000Z K 7 svn:log V 37 Remove an unneeded "a" from comment. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-25T19:48:48.000000Z K 7 svn:log V 2987 Reorganize the interrupt handling code a bit to make a few things cleaner and increase flexibility to allow various different approaches to be tried in the future. - Split struct ithd up into two pieces. struct intr_event holds the list of interrupt handlers associated with interrupt sources. struct intr_thread contains the data relative to an interrupt thread. Currently we still provide a 1:1 relationship of events to threads with the exception that events only have an associated thread if there is at least one threaded interrupt handler attached to the event. This means that on x86 we no longer have 4 bazillion interrupt threads with no handlers. It also means that interrupt events with only INTR_FAST handlers no longer have an associated thread either. - Renamed struct intrhand to struct intr_handler to follow the struct intr_foo naming convention. This did require renaming the powerpc MD struct intr_handler to struct ppc_intr_handler. - INTR_FAST no longer implies INTR_EXCL on all architectures except for powerpc. This means that multiple INTR_FAST handlers can attach to the same interrupt and that INTR_FAST and non-INTR_FAST handlers can attach to the same interrupt. Sharing INTR_FAST handlers may not always be desirable, but having sio(4) and uhci(4) fight over an IRQ isn't fun either. Drivers can always still use INTR_EXCL to ask for an interrupt exclusively. The way this sharing works is that when an interrupt comes in, all the INTR_FAST handlers are executed first, and if any threaded handlers exist, the interrupt thread is scheduled afterwards. This type of layout also makes it possible to investigate using interrupt filters ala OS X where the filter determines whether or not its companion threaded handler should run. - Aside from the INTR_FAST changes above, the impact on MD interrupt code is mostly just 's/ithread/intr_event/'. - A new MI ddb command 'show intrs' walks the list of interrupt events dumping their state. It also has a '/v' verbose switch which dumps info about all of the handlers attached to each event. - We currently don't destroy an interrupt thread when the last threaded handler is removed because it would suck for things like ppbus(8)'s braindead behavior. The code is present, though, it is just under #if 0 for now. - Move the code to actually execute the threaded handlers for an interrrupt event into a separate function so that ithread_loop() becomes more readable. Previously this code was all in the middle of ithread_loop() and indented halfway across the screen. - Made struct intr_thread private to kern_intr.c and replaced td_ithd with a thread private flag TDP_ITHREAD. - In statclock, check curthread against idlethread directly rather than curthread's proc against idlethread's proc. (Not really related to intr changes) Tested on: alpha, amd64, i386, sparc64 Tested on: arm, ia64 (older version of patch by cognet and marcel) END K 10 svn:author V 7 wollman K 8 svn:date V 27 2005-10-25T19:54:25.000000Z K 7 svn:log V 100 We've supported the _POSIX_MONOTONIC_CLOCK option for a long time. Advertise it. MFC after: 1 week END K 10 svn:author V 3 imp K 8 svn:date V 27 2005-10-25T19:59:22.000000Z K 7 svn:log V 102 MFC: 1.26 remove 'string' to avoid shadowing a variable we want to change. Approved by: re@ (scottl) END K 10 svn:author V 3 njl K 8 svn:date V 27 2005-10-25T20:05:44.000000Z K 7 svn:log V 172 MFC: Warn, don't error out when we temporarily can't set the frequency. Accept "maximum" and "minimum" keywords. Set signals in the right place. Approved by: re (scottl) END K 10 svn:author V 3 imp K 8 svn:date V 27 2005-10-25T20:07:47.000000Z K 7 svn:log V 37 MFC: 1.26 Approved by: re@ (scottl) END K 10 svn:author V 3 njl K 8 svn:date V 27 2005-10-25T20:07:51.000000Z K 7 svn:log V 74 MFC: powerd.c rev 1.6.2.3, powerd.8 rev 1.5.2.1 Approved by: re (scottl) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-25T20:10:56.000000Z K 7 svn:log V 97 MFC: Rename 'traceall' to 'alltrace' and add a 'bt' alias for 'trace'. Approved by: re (scottl) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-25T20:11:58.000000Z K 7 svn:log V 97 MF6: Rename 'traceall' to 'alltrace' and add a 'bt' alias for 'trace'. Approved by: re (scottl) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-25T20:14:04.000000Z K 7 svn:log V 139 MFC: Don't panic if a spin lock is initialized that isn't in our static order list. Just warn about it instead. Approved by: re (scottl) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-25T20:17:11.000000Z K 7 svn:log V 139 MF6: Don't panic if a spin lock is initialized that isn't in our static order list. Just warn about it instead. Approved by: re (scottl) END K 10 svn:author V 4 kris K 8 svn:date V 27 2005-10-25T20:33:31.000000Z K 7 svn:log V 104 MFC 1.73: Reflect mpsafety of the underlying filesystem in the nullfs image. Approved by: re (scottl) END K 10 svn:author V 4 kris K 8 svn:date V 27 2005-10-25T20:33:59.000000Z K 7 svn:log V 107 MFC 1.73: Reflect mpsafety of the underlying filesystem in the nullfs image. Approved by: re (scottl) END K 10 svn:author V 4 kris K 8 svn:date V 27 2005-10-25T20:42:06.000000Z K 7 svn:log V 73 MFC r1.650: Default to mpsafevfs=1 on sparc64 Approved by: re (scottl) END K 10 svn:author V 4 kris K 8 svn:date V 27 2005-10-25T20:43:25.000000Z K 7 svn:log V 76 MFC r1.650: Default to mpsafevfs=1 on sparc64 Approved by: re (scottl) END K 10 svn:author V 3 njl K 8 svn:date V 27 2005-10-25T20:51:25.000000Z K 7 svn:log V 99 MFC rev 1.216: Don't disable PCI link devices even when they're !enabled Approved by: re (scottl) END K 10 svn:author V 3 njl K 8 svn:date V 27 2005-10-25T20:51:58.000000Z K 7 svn:log V 103 MFC rev 1.214.2.1: Don't disable PCI link devices even when they're !enabled Approved by: re (scottl) END K 10 svn:author V 3 njl K 8 svn:date V 27 2005-10-25T20:52:44.000000Z K 7 svn:log V 60 MFC rev 1.12: set to 100% on boot Approved by: re (scottl) END K 10 svn:author V 3 njl K 8 svn:date V 27 2005-10-25T20:53:22.000000Z K 7 svn:log V 64 MFC rev 1.11.2.1: set to 100% on boot Approved by: re (scottl) END K 10 svn:author V 2 ru K 8 svn:date V 27 2005-10-25T20:56:12.000000Z K 7 svn:log V 41 Make the cookie constant name canonical. END K 10 svn:author V 2 ru K 8 svn:date V 27 2005-10-25T20:58:30.000000Z K 7 svn:log V 43 Recognize all current standard node types. END K 10 svn:author V 3 njl K 8 svn:date V 27 2005-10-25T21:15:47.000000Z K 7 svn:log V 642 If we're trying to use C2/3 and reads from the register are returning immediately, back off to the next higher Cx sleep state. Some machines with a Via chipset report a valid C3 but a register read doesn't actually halt the CPU. This would cause the machine to appear unresponsive as it repeatedly called cpu_idle() which immediately returned. Causing interrupts (i.e. by pressing the power button) would cause the system to make forward progress, showing that it wasn't actually hung. Also, enable interrupts a little earlier. We don't need them disabled to calculate the delta time for the read. Reported by: silby MFC after: 2 weeks END K 10 svn:author V 6 grehan K 8 svn:date V 27 2005-10-25T21:31:22.000000Z K 7 svn:log V 40 Catch up with interrupt-thread changes. END K 10 svn:author V 5 peter K 8 svn:date V 27 2005-10-26T00:07:25.000000Z K 7 svn:log V 84 MFC: Fix fencepost error that made libkvm fail to read crashdumps. Approved by: re END K 10 svn:author V 6 grehan K 8 svn:date V 27 2005-10-26T00:55:22.000000Z K 7 svn:log V 94 Oops, forgot to add in r1.36 - MFC after: 1 day (it's been in HEAD for quite a while now..) END K 10 svn:author V 6 marcel K 8 svn:date V 27 2005-10-26T01:49:11.000000Z K 7 svn:log V 257 Remove PCI IDs for multiport cards: o Oxford Semiconductor PCI Dual Port Serial o Netmos Nm9845 PCI Bridge with Dual UART Add PCI IDs for single-port cards: o Various SIIG Cyber Serial o Oxford Semiconductor OXCB950 UART Update description as per puc(4). END K 10 svn:author V 6 marcel K 8 svn:date V 27 2005-10-26T02:22:28.000000Z K 7 svn:log V 125 Fix dump of the unread portion of the kernel message buffer. PR: bin/87964 Submitted by: Frank Mayhar frank at exit dot com END K 10 svn:author V 8 takawata K 8 svn:date V 27 2005-10-26T03:24:28.000000Z K 7 svn:log V 36 Add checking for File record magic. END K 10 svn:author V 3 yar K 8 svn:date V 27 2005-10-26T04:12:34.000000Z K 7 svn:log V 89 Document that `reload' is not provided by default yet it can be enabled when applicable. END K 10 svn:author V 3 yar K 8 svn:date V 27 2005-10-26T04:20:18.000000Z K 7 svn:log V 86 Use consistent punctuation and vertical indenting around .Dl's. Wrap a too long line. END K 10 svn:author V 3 yar K 8 svn:date V 27 2005-10-26T04:32:31.000000Z K 7 svn:log V 277 Override $command with $foo_program only if $command has been set in the first place. This should reduce unwanted side-effects in rc.d scripts that don't mean to use $command and rc.subr(8) methods associated with it at all. Discussed with: brooks Reviewed by: -rc (silence) END K 10 svn:author V 3 yar K 8 svn:date V 27 2005-10-26T05:57:35.000000Z K 7 svn:log V 195 Since carp(4) interfaces presently are kinda fake yet possess IP addresses, mark them with LOOPBACK so that routing daemons take them easy for link-state routing protocols. Reviewed by: glebius END K 10 svn:author V 2 ru K 8 svn:date V 27 2005-10-26T06:17:27.000000Z K 7 svn:log V 43 Catch up with new interrupt handling code. END K 10 svn:author V 2 ru K 8 svn:date V 27 2005-10-26T06:44:59.000000Z K 7 svn:log V 43 Catch up with new interrupt handling code. END K 10 svn:author V 5 wpaul K 8 svn:date V 27 2005-10-26T06:52:57.000000Z K 7 svn:log V 1076 Get rid of the timer tracking and reaping code in NdisMInitializeTimer() and ndis_halt_nic(). It's been disabled for some time anyway, and it turns out there's a possible deadlock in NdisMInitializeTimer() when acquiring the miniport block lock to modify the timer list: it's possible for a driver to call NdisMInitializeTimer() when the miniport block lock has already been acquired by an earlier piece of code. You can't acquire the same spinlock twice, so this can deadlock. Also, implement MmMapIoSpace() and MmUnmapIoSpace(), and make NdisMMapIoSpace() and NdisMUnmapIoSpace() use them. There are some drivers that want MmMapIoSpace() and MmUnmapIoSpace() so that they can map arbitrary register spaces not directly associated with their device resources. For example, there's an Atheros driver for a miniPci card (0x168C:0x1014) on the IBM Thinkpad x40 that wants to map some I/O spaces at 0xF00000 and 0xE00000 which are held by the acpi0 device. I don't know what it wants these ranges for, but if it can't map and access them, the MiniportInitialize() method fails. END K 10 svn:author V 7 davidxu K 8 svn:date V 27 2005-10-26T06:55:46.000000Z K 7 svn:log V 510 do umtx_wake at userland thread exit address, so that others userland threads can wait for a thread to exit, and safely assume that the thread has left userland and is no longer using its userland stack, this is necessary for pthread_join when a thread is waiting for another thread to exit which has user customized stack, after pthread_join returns, the userland stack can be reused for other purposes, without this change, the joiner thread has to spin at the address to ensure the thread is really exited. END K 10 svn:author V 5 wpaul K 8 svn:date V 27 2005-10-26T06:56:22.000000Z K 7 svn:log V 126 MFC: implement MmMapIoSpace() and MmUnmapIoSpace(), remove old timer tracking code to fix deadlock in NdisMInitializeTimer(). END K 10 svn:author V 7 davidxu K 8 svn:date V 27 2005-10-26T07:11:43.000000Z K 7 svn:log V 99 Follow the change in kernel, joiner thread just waits at thread id address, let kernel wake it up. END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-10-26T07:18:37.000000Z K 7 svn:log V 177 - Fix leak of struct nlminfo on process exit. - Fix malloc type collision, that made the above problem difficult to understand. Reported by: Vladimir Sharun END K 10 svn:author V 7 davidxu K 8 svn:date V 27 2005-10-26T11:08:32.000000Z K 7 svn:log V 74 Add experiment code to implement POSIX timer's SIGEV_THREAD notification. END K 10 svn:author V 8 keramida K 8 svn:date V 27 2005-10-26T12:07:23.000000Z K 7 svn:log V 85 Typo. PR: docs/88015 Submitted by: Marian Cerny MFC after: 3 days END K 10 svn:author V 3 bde K 8 svn:date V 27 2005-10-26T12:36:18.000000Z K 7 svn:log V 927 Use a better algorithm for reducing the error in __kernel_cos[f](). This supersedes the fix for the old algorithm in rev.1.8 of k_cosf.c. I want this change mainly because it is an optimization. It helps make software cos[f](x) and sin[f](x) faster than the i387 hardware versions for small x. It is also a simplification, and reduces the maximum relative error for cosf() and sinf() on machines like amd64 from about 0.87 ulps to about 0.80 ulps. It was validated for cosf() and sinf() by exhaustive testing. Exhaustive testing is not possible for cos() and sin(), but ucbtest reports a similar reduction for the worst case found by non-exhaustive testing. ucbtest's non-exhaustive testing seems to be good enough to find problems in algorithms but not maximum relative errors when there are spikes. E.g., short runs of it find only 3 ulp error where the i387 hardware cos() has an error of about 2**40 ulps near pi/2. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-26T15:51:05.000000Z K 7 svn:log V 223 Add a swi_remove() function to teardown software interrupt handlers. For now it just calls intr_event_remove_handler(), but at some point it might also be responsible for tearing down interrupt events created via swi_add. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-26T15:52:16.000000Z K 7 svn:log V 144 - Use swi_remove() to teardown swi handlers rather than intr_event_remove_handler(). - Remove tty: prefix from a couple of swi handler names. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-26T17:37:54.000000Z K 7 svn:log V 139 Fix misspelled 'argument' and try to make the sentence into English. Submitted by: Andre Guibert de Bruet andy at siliconlandmark dot com END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-26T17:39:35.000000Z K 7 svn:log V 98 Spell 'argument' correctly. Submitted by: Andre Guibert de Bruet andy at siliconlandmark dot com END K 10 svn:author V 5 wpaul K 8 svn:date V 27 2005-10-26T18:46:27.000000Z K 7 svn:log V 1312 Clean up and apply the fix for PR 83477. The calculation for locating the start of the section headers has to take into account the fact that the image_nt_header is really variable sized. It happens that the existing calculation is correct for _most_ production binaries produced by the Windows DDK, but if we get a binary with oddball offsets, the PE loader could crash. Changes from the supplied patch are: - We don't really need to use the IMAGE_SIZEOF_NT_HEADER() macro when computing how much of the header to return to callers of pe_get_optional_header(). While it's important to take the variable size of the header into account in other calculations, we never actually look at anything outside the non-variable portion of the header. This saves callers from having to allocate a variable sized buffer off the heap (I purposely tried to avoid using malloc() in subr_pe.c to make it easier to compile in both the -D_KERNEL and !-D_KERNEL case), and since we're copying into a buffer on the stack, we always have to copy the same amount of data or else we'll trash the stack something fierce. - We need to get offsetof() in the !-D_KERNEL case. - ndiscvt.c needs the IMAGE_FIRST_SECTION() macro too, since it does a little bit of section pre-processing. PR: kern/83477 END K 10 svn:author V 5 wpaul K 8 svn:date V 27 2005-10-26T18:48:57.000000Z K 7 svn:log V 49 MFC: merge the fix for PR 83477. PR: kern/83477 END K 10 svn:author V 3 pjd K 8 svn:date V 27 2005-10-26T19:44:17.000000Z K 7 svn:log V 170 Change u_int64_t to uintmax_t and use %ju, so it compiles on 64bit archs. I changed every u_int64_t to uintmax_t, as we should use eventually uint64_t anyway those days. END K 10 svn:author V 3 pjd K 8 svn:date V 27 2005-10-26T19:57:04.000000Z K 7 svn:log V 646 Call if_clone_detach() after vlan_clone_destroy() loop, as if_clone_detach() is freeing ifc_units field, which is used in vlan_clone_destroy()->ifc_free_unit(). This fixes panic: ifc_free_unit: bit is already cleared The fix introduces a small race in that a new interface could be created between the vlan_clone_destroy() loop and the call to if_clone_detach(). The race is probably too hard to trigger to be worry about it. This is not an MFC, because better fix was committed to HEAD. We decided to use this fix, as it is a low-risk fix, which fits just fine for 6.0 at this stage. Discussed with: brooks, thompsa Approved by: re (scottl) END K 10 svn:author V 8 keramida K 8 svn:date V 27 2005-10-26T20:19:39.000000Z K 7 svn:log V 185 Note that the jail setup example is meant to be fed to sh(1), not csh(1). PR: docs/87351 Submitted by: "Eli K. Breen" Approved by: simon, brooks MFC after: 3 days END K 10 svn:author V 7 delphij K 8 svn:date V 27 2005-10-26T20:21:23.000000Z K 7 svn:log V 244 MFC (by alc) changesets that addresses several race conditions that can cause a kernel compiled with ZERO_COPY_SOCKETS to panic under certain circumstances: sys/kern/uipc_cow.c: 1.24 - 1.26 sys/vm/vm_object.c: 1.351 Approved by: re (scottl) END K 10 svn:author V 5 wpaul K 8 svn:date V 27 2005-10-26T20:21:45.000000Z K 7 svn:log V 144 Minor nit: in ntoskrnl_finddev(), only free the 'children' device_t array if device_find_children() actually returned a non-NULL array pointer. END K 10 svn:author V 7 delphij K 8 svn:date V 27 2005-10-26T20:22:00.000000Z K 7 svn:log V 251 MFRELENG_6 (by alc) changesets that addresses several race conditions that can cause a kernel compiled with ZERO_COPY_SOCKETS to panic under certain circumstances: sys/kern/uipc_cow.c: 1.24 - 1.26 sys/vm/vm_object.c: 1.351 Approved by: re (scottl) END K 10 svn:author V 5 wpaul K 8 svn:date V 27 2005-10-26T20:23:29.000000Z K 7 svn:log V 94 MFC: only free child device array if device_get_children() returned a non-NULL array pointer. END K 10 svn:author V 4 jkim K 8 svn:date V 27 2005-10-26T21:13:30.000000Z K 7 svn:log V 87 MFC: Fix a longstanding buglet in bz-prefixed grep(1). PR: gnu/72200 Approved by: re END K 10 svn:author V 4 jkim K 8 svn:date V 27 2005-10-26T21:15:51.000000Z K 7 svn:log V 94 MFRELENG_6: Fix a longstanding buglet in bz-prefixed grep(1). PR: gnu/72200 Approved by: re END K 10 svn:author V 5 ariff K 8 svn:date V 27 2005-10-26T21:17:40.000000Z K 7 svn:log V 203 Track every mixer device closing for proper refcount calculation. This solve device busy and panic issues during sound module unloading. Approved by: re (scottl), netchild (mentor) Reviewed by: andreas END K 10 svn:author V 5 ariff K 8 svn:date V 27 2005-10-26T21:18:09.000000Z K 7 svn:log V 282 Rearrange sound unregister procedure to detect mixer busy state and destroy it before anything else. This solve panic issue during sound module unloading if the sound / mixer device being held by external processes. Approved by: re (scottl), netchild (mentor) Reviewed by: andreas END K 10 svn:author V 5 ariff K 8 svn:date V 27 2005-10-26T21:18:41.000000Z K 7 svn:log V 221 Virtual channel default speed should be based on soundcard capabilities. This solve various mysterious slowdown especially for fixed-rate ac97 soundcards. Approved by: re (scottl), netchild (mentor) Reviewed by: andreas END K 10 svn:author V 4 csjp K 8 svn:date V 27 2005-10-26T21:58:42.000000Z K 7 svn:log V 94 MFC notes about ucred based rules and debug.mpsafenet requirements. Approved by: re (scottl) END K 10 svn:author V 5 peter K 8 svn:date V 27 2005-10-26T22:07:03.000000Z K 7 svn:log V 61 MFC: 1.2: don't set segment registers yet. Approved by: re END K 10 svn:author V 5 peter K 8 svn:date V 27 2005-10-26T22:16:52.000000Z K 7 svn:log V 115 Change PHYSMAP_SIZE to allow for more memory segments. The old value was too low for certain Dell amd64 machines. END K 10 svn:author V 5 peter K 8 svn:date V 27 2005-10-26T22:19:51.000000Z K 7 svn:log V 191 There is no 'freebsd3_' prefix for COMPAT_43 syscalls. Those are all bundled under MCOMPAT and have an 'o' prefix. Adjust as appropriate. This re-enables compiling without COMPAT_43 again. END K 10 svn:author V 5 peter K 8 svn:date V 27 2005-10-26T22:21:03.000000Z K 7 svn:log V 57 Regenerate (with the correct #ifdef COMPAT_43 tests now) END K 10 svn:author V 5 peter K 8 svn:date V 27 2005-10-26T22:23:52.000000Z K 7 svn:log V 254 Make fdcontrol work again. It has been broken for a while. It tries to set the floppy controller parameters, but that requires that the device node be open in O_RDWR mode now. I think it is broken in 6.0 as well. This line looks like a stray anyway. END K 10 svn:author V 5 peter K 8 svn:date V 27 2005-10-26T22:28:38.000000Z K 7 svn:log V 332 fdread cannot work without opening the /dev/fd node in O_RDWR mode now, because it sets the floppy controller parameters, which requires O_RDWR. Specifically, the FD_SOPTS ioctl requires this, and the code errors out and aborts if it can't do it. Among other things, it is changing the FDOPT_NOERRLOG flag. Broken in 6.0 as well. END K 10 svn:author V 5 peter K 8 svn:date V 27 2005-10-26T22:32:30.000000Z K 7 svn:log V 241 MFi386: Various apic fixes and tweaks * Don't recursively panic if we've already paniced and the local apic is now stuck. * Add hw.apic.* tunables/sysctls for extint controls * Change "lapic%d timer" to "cpu%d timer" intname to match i386 END K 10 svn:author V 5 peter K 8 svn:date V 27 2005-10-26T22:40:07.000000Z K 7 svn:log V 355 Commit something we found useful at work at one point. Add sysctls for debug.kdb.panic and debug.kdb.trap alongside the existing debug.kdb.enter sysctl. 'panic' causes a panic, and 'trap' causes a page fault. We used these to ensure that crash dumps succeed from those two common failure modes. This avoids the need for creating a 'panic' kld module. END K 10 svn:author V 4 emax K 8 svn:date V 27 2005-10-26T23:13:51.000000Z K 7 svn:log V 95 Do not manually allocate/free device's softc structure. Pointed by: jhb Reviewed by: imp, jhb END K 10 svn:author V 7 davidxu K 8 svn:date V 27 2005-10-27T03:09:20.000000Z K 7 svn:log V 190 Disconnect libc_r from buildworld, it is still kept in the tree to provide some baseline references, but users are encouraged to use libpthread or libthr in real world. Discussed on: arch@ END K 10 svn:author V 2 ps K 8 svn:date V 27 2005-10-27T04:26:35.000000Z K 7 svn:log V 69 Allow 32bit get/setsockopt with SO_SNDTIMEO or SO_RECVTIMEO to work. END K 10 svn:author V 7 davidxu K 8 svn:date V 27 2005-10-27T10:21:23.000000Z K 7 svn:log V 46 Link libthr to libpthread on Alpha and Sparc. END K 10 svn:author V 3 suz K 8 svn:date V 27 2005-10-27T14:04:03.000000Z K 7 svn:log V 135 MFC rev 1.34 fixed a kernel crash when IPv6 PIM-SM routing is enabled and a PIM register message is received Approved by: re(scottl) END K 10 svn:author V 2 ru K 8 svn:date V 27 2005-10-27T14:24:45.000000Z K 7 svn:log V 288 Installing debug modules was a bad idea -- I bogusly assumed that our kernel linker will only load PT_LOAD segments, apparently not. Instead, produce .dbg objects from .debug objects, and install them together with non-debug objects, as described in objcopy(1). Original code by: obrien END K 10 svn:author V 2 ru K 8 svn:date V 27 2005-10-27T14:33:08.000000Z K 7 svn:log V 153 Use ${S} to pass ${SYSDIR} to ports. This makes PORTS_MODULES feature work when compiling a kernel via "make buildkernel". Noticed and tested by: nork END K 10 svn:author V 6 obrien K 8 svn:date V 27 2005-10-27T15:39:19.000000Z K 7 svn:log V 75 Clear pending_txs when not "RUNNING". Submitted by: Q END K 10 svn:author V 4 jkim K 8 svn:date V 27 2005-10-27T16:26:27.000000Z K 7 svn:log V 140 MFC: Fix 'grep -Fw' for encodings other than UTF-8 (RH bug #161700). PR: 87969 Approved by: re (scottl) Obtained from: Fedora (Tim Waugh) END K 10 svn:author V 4 jkim K 8 svn:date V 27 2005-10-27T16:29:02.000000Z K 7 svn:log V 147 MFRELENG_6: Fix 'grep -Fw' for encodings other than UTF-8 (RH bug #161700). PR: 87969 Approved by: re (scottl) Obtained from: Fedora (Tim Waugh) END K 10 svn:author V 3 sos K 8 svn:date V 27 2005-10-27T16:32:39.000000Z K 7 svn:log V 201 Enclose the delayed attach in Giant so we dont loose the race with other drivers trying to attach ATA devices like pccard. Dont clear the delayed flag before we are acutally finished. Spotted by: imp END K 10 svn:author V 3 sos K 8 svn:date V 27 2005-10-27T16:41:39.000000Z K 7 svn:log V 219 MFC: Enclose the delayed attach in Giant so we dont loose the race with other drivers trying to attach ATA devices like pccard. Dont clear the delayed flag before we are acutally finished. Approved by: re@ (scottl) END K 10 svn:author V 5 wpaul K 8 svn:date V 27 2005-10-27T16:41:49.000000Z K 7 svn:log V 219 MFC: add 1 microsecond delay to pci_add_children() bus enumeration loop to pacify certain PCI bridges ('fixes' problem with USB controllers not being probed on some Sun AMD8111 machines, notably mine). Approved by: re END K 10 svn:author V 5 wpaul K 8 svn:date V 27 2005-10-27T16:44:06.000000Z K 7 svn:log V 89 MFC: add 1 microsecond delay to pci_add_children() bus enumeration loop Approved by: re END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-27T16:47:28.000000Z K 7 svn:log V 839 Revert most of revision 1.235 and fix the problem a different way. We can't acquire an sx lock in ttyinfo() because ttyinfo() can be called from interrupt handlers (such as atkbd_intr()). Instead, go back to locking the process group while we pick a thread to display information for and hold that lock until after we drop sched_lock to make sure the process doesn't exit out from under us. sched_lock ensures that the specific thread from that process doesn't go away. To protect against the process exiting after we drop the proc lock but before we dereference it to lookup the pid and p_comm in the call to ttyprintf(), we now copy the pid and p_comm to local variables while holding the proc lock. This problem was found by the recently added TD_NO_SLEEPING assertions for interrupt handlers. Tested by: emaste MFC after: 1 week END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-27T16:59:02.000000Z K 7 svn:log V 172 Update example libmap.conf file: - Remove pointless examples that map libpthread to itself. - Update shared libraries for 6.0 bumps. Reported by: rwatson MFC after: 1 day END K 10 svn:author V 5 wpaul K 8 svn:date V 27 2005-10-27T17:08:57.000000Z K 7 svn:log V 124 MFC: synchronize the NDISulator with the head and RELENG_5 branches, add -D ndis support to wpa_supplicant Approved by: re END K 10 svn:author V 7 cvs2svn K 8 svn:date V 27 2005-10-27T17:08:58.000000Z K 7 svn:log V 70 This commit was manufactured by cvs2svn to create branch 'RELENG_6_1'. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-27T17:13:23.000000Z K 7 svn:log V 563 Optionally include a DEFAULTS config file if it is present in the current directory before the specified config file. This is implemented by opening DEFAULTS as stdin if it exists, and if so resetting stdin to the actual config file when DEFAULTS is fully parsed via yywrap(). In short, this lets us create DEFAULTS kernel configs in /sys//conf that can enable certain options or devices by default and allow users to disable them via 'nooptions' or 'nodevice' rather than having to create kludge NO_FOO options. Requested by: scottl Reviewed by: scottl END K 10 svn:author V 5 wpaul K 8 svn:date V 27 2005-10-27T17:15:24.000000Z K 7 svn:log V 16 Approved by: re END K 10 svn:author V 5 wpaul K 8 svn:date V 27 2005-10-27T17:15:47.000000Z K 7 svn:log V 131 MFC: add the driver_ndis support files to the wpa_supplicant distribution for -D ndis support with the NDISulator Approved by: re END K 10 svn:author V 6 emaste K 8 svn:date V 27 2005-10-27T17:22:47.000000Z K 7 svn:log V 293 In watchdog_config enable the software watchdog iff the WD_ACTIVE flag is set. When watchdogd(1) is terminated intentionally it clears the bit, which should then disable it in the kernel. PR: kern/74386 Submitted by: Alex Hoff Approved by: phk, rwatson (mentor) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-27T17:34:35.000000Z K 7 svn:log V 281 Create a default kernel config for i386 and move 'device isa' and 'device npx' (both of which aren't really optional right now) and 'device io' and 'device mem' (to preserve POLA for 4.x users upgrading to 6.0) from GENERIC into DEFAULTS. Requested by: scottl Reviewed by: scottl END K 10 svn:author V 7 delphij K 8 svn:date V 27 2005-10-27T17:35:48.000000Z K 7 svn:log V 99 MFen: 1.883.2.7 -> 1.883.2.8 MT6 candidate Obtained from: The FreeBSD Simplified Chinese Project END K 10 svn:author V 2 ru K 8 svn:date V 27 2005-10-27T17:39:03.000000Z K 7 svn:log V 220 Rename the .dbg extension to .symbols, which matches "symbol-file" gdb(1) command better, though I must admit it's confusing: these files have not only [debugging] symbols, but much more than that. Requested by: obrien END K 10 svn:author V 6 scottl K 8 svn:date V 27 2005-10-27T17:48:57.000000Z K 7 svn:log V 138 MFC: Sync with HEAD. This fixes an out-of-bounds access that was caught by INVARIANTS, and it synchronizes the license. Approved by: re END K 10 svn:author V 7 delphij K 8 svn:date V 27 2005-10-27T17:53:31.000000Z K 7 svn:log V 89 MFC 1.9: Update to sync up with latest RELENG_6 documentation. Approved by: re (scottl) END K 10 svn:author V 7 delphij K 8 svn:date V 27 2005-10-27T17:54:31.000000Z K 7 svn:log V 100 MFRELENG_6 1.5.2.3: Update to sync up with latest RELENG_6 documentation. Approved by: re (scottl) END K 10 svn:author V 6 murray K 8 svn:date V 27 2005-10-27T18:34:51.000000Z K 7 svn:log V 136 Look for packages in packages-6.0-release or packages-6-stable, as appropriate. Approved by: re (scottl@) Reviewed by: portmgr (kris@) END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-10-27T18:35:19.000000Z K 7 svn:log V 216 MFC: - Fix leak of struct nlminfo on process exit. - Fix malloc type collision, that made the above problem difficult to understand. Reported by: Vladimir Sharun Approved by: re (scottl) END K 10 svn:author V 6 murray K 8 svn:date V 27 2005-10-27T18:36:13.000000Z K 7 svn:log V 133 MFC: look for packages in the packages-6-release directory when appropriate. Reviewed by: portmgr (kris@) Approved by: re (scottl@) END K 10 svn:author V 6 murray K 8 svn:date V 27 2005-10-27T18:38:48.000000Z K 7 svn:log V 85 MFC: Use packages-6-release directory for packages, as appropriate. Approved by: re END K 10 svn:author V 6 obrien K 8 svn:date V 27 2005-10-27T18:40:56.000000Z K 7 svn:log V 70 Remove atpic as we've changed to using the lapic timer vs. using irq0 END K 10 svn:author V 5 peter K 8 svn:date V 27 2005-10-27T18:54:43.000000Z K 7 svn:log V 147 MFi386: bring over DEFAULTS (repocopy) and adapt. While there isn't a 4.x->6.x amd64 upgrade path, the config files are kept in approximate sync. END K 10 svn:author V 6 scottl K 8 svn:date V 27 2005-10-27T18:58:37.000000Z K 7 svn:log V 94 MFC: Sync the mpt driver to pick up license refinement and INVARIANTS error. Approved by: re END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-10-27T19:02:34.000000Z K 7 svn:log V 73 Keep locks consistent before goto. Reported by: pho Reviewed by: mohans END K 10 svn:author V 6 obrien K 8 svn:date V 27 2005-10-27T19:07:35.000000Z K 7 svn:log V 78 Remove information on nVidia nForce3 Pro-150 chipset to match GENERIC change. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-27T19:27:55.000000Z K 7 svn:log V 48 Bump config(8) version for the DEFAULTS change. END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2005-10-27T19:41:43.000000Z K 7 svn:log V 336 MFC if_bridge r1.30 > If we have been called from ether_ifdetach() then do not try and clear the > promisc flag from the member interface, this is a no-op anyway since the > interface is disappearing. The driver may have already released > its resources such as miibus and this is likely to panic the kernel. Approved by: re (scottl) END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2005-10-27T19:43:07.000000Z K 7 svn:log V 339 MFC if_bridge r1.30 > If we have been called from ether_ifdetach() then do not try and clear the > promisc flag from the member interface, this is a no-op anyway since the > interface is disappearing. The driver may have already released > its resources such as miibus and this is likely to panic the kernel. Approved by: re (scottl) END K 10 svn:author V 4 joel K 8 svn:date V 27 2005-10-27T20:39:37.000000Z K 7 svn:log V 110 MFC r1.1334: snd_ess needs snd_sbc, so add a note about that. Requested by: delphij Approved by: re (scottl) END K 10 svn:author V 4 joel K 8 svn:date V 27 2005-10-27T20:47:03.000000Z K 7 svn:log V 91 MFC r1.59: Add support for ATI IXP[234]00. Requested by: delphij Approved by: re (scottl) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-27T20:48:05.000000Z K 7 svn:log V 157 Remove old ACPICA files from the INTEL vendor branch. They were removed from HEAD about 4 years ago when we started flattening out the ACPICA distribution. END K 10 svn:author V 8 kensmith K 8 svn:date V 27 2005-10-27T20:55:29.000000Z K 7 svn:log V 421 Temporary hack to get the hme network interface on Sun E250 servers to work. The code comment describes the issue, basically an as yet not totally understood interrupt routing problem. This hack is being done to RELENG_6 so that the hme interface works on E250's for the release, but is not being done in HEAD so more work on the interrupt routing issue can be done. Requested by: marius Approved by: re (scottl) END K 10 svn:author V 8 kensmith K 8 svn:date V 27 2005-10-27T20:58:50.000000Z K 7 svn:log V 453 MFC of Rev. 1.6.2.1: > Temporary hack to get the hme network interface on Sun E250 servers to > work. The code comment describes the issue, basically an as yet not > totally understood interrupt routing problem. This hack is being done > to RELENG_6 so that the hme interface works on E250's for the release, > but is not being done in HEAD so more work on the interrupt routing > issue can be done. > > Requested by: marius Approved by: re (scottl) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-27T21:08:12.000000Z K 7 svn:log V 144 Drop the driver lock around atm_input() analogous to all the ethernet drivers dropping the driver lock around ether_input(). Silence by: harti END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-27T21:16:17.000000Z K 7 svn:log V 632 - Add locking and mark MPSAFE. The driver had a mutex in the softc and even initialized it, but it never used it. - Use callout_*() to manage the callout. - Use m_devget() to copy data out of the rx buffers rather than doing it all by hand. - Use m_getcl() to allocate mbuf clusters rather than doing it all by hand. - Don't free the software descriptor for a rx ring entry if we can't allocate an mbuf cluster for it. We left a dangling pointer and never reallocated the entry anyway. OpenBSD's code (from which this was derived) has the same bug. Tested by: NO ONE (despite repeated requests) Reviewed by: wpaul (5) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-27T21:18:37.000000Z K 7 svn:log V 269 - Use if_printf() and device_printf() and remove vr_unit from the softc. I had to initialize the ifnet a bit earlier in attach so that the if_printf()'s in vr_reset() didn't explode with a page fault. - Use M_ZERO with contigmalloc() rather than an explicit bzero. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-27T21:22:58.000000Z K 7 svn:log V 518 Fixup locking and mark MPSAFE. - Add locked variants of init() and start(). - Use callout_*() to manage callout. - Test IFF_DRV_RUNNING rather than IFF_UP in wb_intr() to see if we are still active when an interrupt comes in. I couldn't find any of these cards anywhere to test on myself, and google turns up references to FreeBSD and OpenBSD manpages for this driver when trying to locate a card that way. I'm not sure anyone actually uses these cards with FreeBSD. Tested by: NO ONE (despite repeated requests) END K 10 svn:author V 7 rodrigc K 8 svn:date V 27 2005-10-27T23:57:55.000000Z K 7 svn:log V 246 Submitted following patch to FSF GCC: * freebsd-spec.h (FBSD_TARGET_OS_CPP_BUILTINS): Use builtin_define_with_int_value() instead of adding a new check for every new major FreeBSD version. Motivated by: simon Discussed with: obrien, kan END K 10 svn:author V 6 grehan K 8 svn:date V 27 2005-10-28T00:58:43.000000Z K 7 svn:log V 297 MFC r1.36 > The BCM5401 dspcode load on media changes also applies > to the 100/1000 BCM5400 phy. This fixes the problem with > the GEM port not syncing up on Sawtooth G4's. > > Obtained from: NetBSD > Reported by: Ben Rosengart Approved by: re (scottl) END K 10 svn:author V 3 ups K 8 svn:date V 27 2005-10-28T01:03:26.000000Z K 7 svn:log V 222 MFC vfs_cluster.c Revision 1.168 Only set B_RAM (Read ahead mark) on an incore buffers if we can lock it. This fixes a race condition caused by the unlocked write access to the b_flags field. Approved by: re (scottl) END K 10 svn:author V 6 grehan K 8 svn:date V 27 2005-10-28T01:07:54.000000Z K 7 svn:log V 324 MF RELENG_6 r1.34.2.1 > The BCM5401 dspcode load on media changes also applies > to the 100/1000 BCM5400 phy. This fixes the problem with > the GEM port not syncing up on Sawtooth G4's. > > Obtained from: NetBSD > Reported by: Ben Rosengart Approved by: re (scottl) END K 10 svn:author V 5 wpaul K 8 svn:date V 27 2005-10-28T02:17:40.000000Z K 7 svn:log V 489 Remove forgotten, no longer needed WB_UNLOCK() from the end wb_ioctl(). With this change, the driver tests good (at least on i386): wb0: port 0xb800-0xb87f mem 0xe6800000-0xe680007f irq 12 at device 10.0 on pci0 miibus1: on wb0 amphy0: on miibus1 amphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto wb0: Ethernet address: 00:00:e8:18:2a:02 wb0: link state changed to DOWN wb0: link state changed to UP END K 10 svn:author V 7 delphij K 8 svn:date V 27 2005-10-28T02:58:50.000000Z K 7 svn:log V 605 Some changes to reflect 6.0-STABLE/RELEASE: - Mention that this is -STABLE - Mention that only 5.3+ is supported for source upgrade - When just building a kernel we assume that the user got a system that is fresh enough - The in-place update is supported for 5.x-STABLE and 6.x-STABLE - Since we don't support systems that is prior to 5.3, don't mention about mergemaster three years before. - Since source upgrade from 4.x is no longer supported, we are supposed that COMPAT_FREEBSD5X is better suited for the source upgrade - At the end, call this a -STABLE Approved by: re (scottl) END K 10 svn:author V 7 delphij K 8 svn:date V 27 2005-10-28T02:59:51.000000Z K 7 svn:log V 92 MFRELENG_6 1.416.2.4: Some changes to reflect 6.0-STABLE/RELEASE. Approved by: re (scottl) END K 10 svn:author V 3 imp K 8 svn:date V 27 2005-10-28T05:29:41.000000Z K 7 svn:log V 55 Use PCIR_BARS rather than CARDBUS_BASE0_REG Style nit. END K 10 svn:author V 3 imp K 8 svn:date V 27 2005-10-28T05:30:47.000000Z K 7 svn:log V 66 Simplify code a little, prefer PCI?_FOO registers where possible. END K 10 svn:author V 3 imp K 8 svn:date V 27 2005-10-28T05:31:11.000000Z K 7 svn:log V 30 Remove now redundant defines. END K 10 svn:author V 3 imp K 8 svn:date V 27 2005-10-28T05:55:52.000000Z K 7 svn:log V 73 Eliminate even more duplication, and move some definitions into pcireg.h END K 10 svn:author V 3 imp K 8 svn:date V 27 2005-10-28T05:56:50.000000Z K 7 svn:log V 23 Minor style(9) nitage. END K 10 svn:author V 3 imp K 8 svn:date V 27 2005-10-28T05:57:10.000000Z K 7 svn:log V 70 Some cardbus CIS definitions moved here since they are PCI registers. END K 10 svn:author V 3 imp K 8 svn:date V 27 2005-10-28T06:02:51.000000Z K 7 svn:log V 28 Add PCIM_BIOS_ADDR_MASK too END K 10 svn:author V 3 imp K 8 svn:date V 27 2005-10-28T06:03:53.000000Z K 7 svn:log V 47 Cardbus has only 1 slot, so simplify a little. END K 10 svn:author V 6 marcel K 8 svn:date V 27 2005-10-28T06:24:09.000000Z K 7 svn:log V 128 Allow uart_bus_probe() to return the predefined BUS_PROBE_* constants. In this case: just return what uart_bus_probe() returns. END K 10 svn:author V 6 marcel K 8 svn:date V 27 2005-10-28T06:27:53.000000Z K 7 svn:log V 255 o Style(9) nits o Fix typo in comment o s/-100/BUS_PROBE_GENERIC/ o s/err/error/ for consistency o Remove non-applicable comment o Allow uart_bus_probe() to return the predefined BUS_PROBE_* contants. In this case: explicitly test for error > 0. END K 10 svn:author V 6 marcel K 8 svn:date V 27 2005-10-28T06:30:39.000000Z K 7 svn:log V 75 In uart_bus_probe() return BUS_PROBE_DEFAULT when the probe is successful. END K 10 svn:author V 3 ade K 8 svn:date V 27 2005-10-28T06:49:49.000000Z K 7 svn:log V 249 MFC: Specifically panic() in the case where pmap_insert_entry() fails to get a new pv under high system load where the available pv entries have been exhausted before the pagedaemon has a chance to wake up to reclaim some. Approved by: re (scottl) END K 10 svn:author V 3 ade K 8 svn:date V 27 2005-10-28T06:50:36.000000Z K 7 svn:log V 223 MFC: Specifically panic() in the case where pmap_insert_entry() fails to get a new pv under high system load where the available pv entries have been exhausted before the pagedaemon has a chance to wake up to reclaim some. END K 10 svn:author V 7 stefanf K 8 svn:date V 27 2005-10-28T10:45:19.000000Z K 7 svn:log V 496 Protect malloc, realloc and free calls with INT{ON,OFF} directly in chkalloc, ckrealloc and ckfree (added), respectively. sh jumps out of the signal handler using longjmp which is obviously a bad idea during malloc calls. Note: I think there is still a small race here because volatile sig_atomic_t only guarantees atomic reads and writes while we're doing increments and decrements. Protect a setmode call with INT{ON,OFF} as it calls malloc internally. PR: 45478 Patch from: Nate Eldredge END K 10 svn:author V 3 bde K 8 svn:date V 27 2005-10-28T13:36:58.000000Z K 7 svn:log V 988 Use fairly optimal minimax polynomials for __kernel_cosf() and __kernel_sinf(). The old ones were the double-precision polynomials with coefficients truncated to float. Truncation is not a good way to convert minimax polynomials to lower precision. Optimize for efficiency and use the lowest-degree polynomials that give a relative error of less than 1 ulp -- degree 8 instead of 14 for cosf and degree 9 instead of 13 for sinf. For sinf, the degree 8 polynomial happens to be 6 times more accurate than the old degree 14 one, but this only gives a tiny amount of extra accuracy in results -- we just need to use a a degree high enough to give a polynomial whose relative accuracy in infinite precision (but with float coefficients) is a small fraction of a float ulp (fdlibm generally uses 1/32 for the small fraction, and the fraction for our degree 8 polynomial is about 1/600). The maximum relative errors for cosf() and sinf() are now 0.7719 ulps and 0.7969 ulps, respectively. END K 10 svn:author V 7 stefanf K 8 svn:date V 27 2005-10-28T14:02:42.000000Z K 7 svn:log V 202 Add test cases for the command built-in, including its -v and -V options which I'll commit real soon. Count the number of found test cases instead of hard-coding them. Allow an arbitrary exit status. END K 10 svn:author V 2 ru K 8 svn:date V 27 2005-10-28T14:15:54.000000Z K 7 svn:log V 70 Belatedly document the NGM_KSOCKET_ACCEPT ("accept") control message. END K 10 svn:author V 3 suz K 8 svn:date V 27 2005-10-28T14:38:56.000000Z K 7 svn:log V 93 MFC Rev 1.60 perform NUD on an IPv6-aware point-to-point interface Approved by: re(scottl) END K 10 svn:author V 2 ru K 8 svn:date V 27 2005-10-28T14:41:28.000000Z K 7 svn:log V 45 Use variable-sized arrays where appropriate. END K 10 svn:author V 4 nyan K 8 svn:date V 27 2005-10-28T15:30:51.000000Z K 7 svn:log V 86 Move the isa, npx, mem and io devices and the PC98 option from GENERIC into DEFAULTS. END K 10 svn:author V 7 cvs2svn K 8 svn:date V 27 2005-10-28T15:30:52.000000Z K 7 svn:log V 68 This commit was manufactured by cvs2svn to create branch 'RELENG_6'. END K 10 svn:author V 7 cvs2svn K 8 svn:date V 27 2005-10-28T15:30:53.000000Z K 7 svn:log V 70 This commit was manufactured by cvs2svn to create branch 'RELENG_6_0'. END K 10 svn:author V 6 scottl K 8 svn:date V 27 2005-10-28T15:48:09.000000Z K 7 svn:log V 70 MFC: Fix regression with Dell PERC3/ROMB controllers Approved by: re END K 10 svn:author V 5 joerg K 8 svn:date V 27 2005-10-28T15:58:19.000000Z K 7 svn:log V 343 Finally complete some work on generalizing the PCF8584-based I2C drivers I started quite some time before. Retire the old i386-only pcf driver, and activate the new general driver that has been sitting in the tree already for quite some time. Build the i2c modules for sparc64 architectures as well (where I've been developing all this on). END K 10 svn:author V 3 yar K 8 svn:date V 27 2005-10-28T16:07:52.000000Z K 7 svn:log V 290 Transforming "ppp-user" into just "ppp", step 1: The rcorder(8) condition PROVIDE'd by the script and REQUIRE'd by the others becomes "ppp". The ultimate goal of the transformation is to reduce confusion resulting from the fact that $name has been "ppp" already. Discussed with: pjd, -rc END K 10 svn:author V 3 yar K 8 svn:date V 27 2005-10-28T16:10:56.000000Z K 7 svn:log V 117 Use: command="/path/to/${name}" since it's applicable here. It's the current style of rc.d. Pointed out by: pjd END K 10 svn:author V 6 scottl K 8 svn:date V 27 2005-10-28T16:31:34.000000Z K 7 svn:log V 64 MFC: Fix regression in Dell PERC/ROMB systems. Approved by: re END K 10 svn:author V 3 yar K 8 svn:date V 27 2005-10-28T16:55:38.000000Z K 7 svn:log V 128 Use ${name} in pathnames where appropriate. The sendmail script already was on this way, but it didn't reach the end of it yet. END K 10 svn:author V 7 stefanf K 8 svn:date V 27 2005-10-28T18:37:09.000000Z K 7 svn:log V 200 Add the POSIX options -v and -V to the 'command' builtin. Both describe the type of their argument, if it is a shell function, an alias, a builtin, etc. -V is more verbose than -v. PR: 77259, 84539 END K 10 svn:author V 8 cracauer K 8 svn:date V 27 2005-10-28T18:39:00.000000Z K 7 svn:log V 206 Fix this: kern/87959 cracauer ext2fs: no cp(1) possible, mmap returns EINVAL ext2fs was missing vnode_create_vobject. (Reisefs probably has the same problem but I want to get this in quick for 6-release) END K 10 svn:author V 8 cracauer K 8 svn:date V 27 2005-10-28T18:44:26.000000Z K 7 svn:log V 286 Fix this: kern/87959 cracauer ext2fs: no cp(1) possible, mmap returns EINVAL ext2fs was missing vnode_create_vobject. (Reisefs probably has the same problem but I want to get this in quick for 6-release) releng-6 branches commits approved by Scott Long. Approved by: re END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-28T19:02:22.000000Z K 7 svn:log V 52 MFC: Add COMPAT_FREEBSD5. Approved by: re (scottl) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-28T19:04:03.000000Z K 7 svn:log V 196 MFC: Sync config(8) with HEAD: - Add -V option to dump version. - Implicitly include DEFAULTS if it exists in the current directory before the specified kernel config. Approved by: re (scottl) END K 10 svn:author V 2 ru K 8 svn:date V 27 2005-10-28T19:05:44.000000Z K 7 svn:log V 32 Document NGM_IFACE_GET_IFINDEX. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-28T19:20:00.000000Z K 7 svn:log V 52 MF6: Add COMPAT_FREEBSD5. Approved by: re (scottl) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-28T19:20:53.000000Z K 7 svn:log V 67 MF6: config -V and support for DEFAULTS. Approved by: re (scottl) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-28T19:21:27.000000Z K 7 svn:log V 77 MFC: Add DEFAULTS files for amd64, i386, and pc98. Approved by: re (scottl) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-28T19:22:41.000000Z K 7 svn:log V 77 MF6: Add DEFAULTS files for amd64, i386, and pc98. Approved by: re (scottl) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-28T19:24:13.000000Z K 7 svn:log V 78 MFC: Sync with head updating examples to match 6.x. Approved by: re (scottl) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-10-28T19:24:41.000000Z K 7 svn:log V 73 MF6: Sync with HEAD updating examples for 6.x. Approved by: re (scottl) END K 10 svn:author V 3 pjd K 8 svn:date V 27 2005-10-28T20:25:02.000000Z K 7 svn:log V 228 Fix possible live-lock under heavy load where we can't allocate more memory for request. I was sure graid3 should handle such situations well, but green@ reported it is not and we want to fix it before 6.0. Submitted by: green END K 10 svn:author V 3 pjd K 8 svn:date V 27 2005-10-28T20:27:53.000000Z K 7 svn:log V 253 Fix possible live-lock under heavy load where we can't allocate more memory for request. I was sure graid3 should handle such situations well, but green@ reported it is not and we want to fix it before 6.0. Submitted by: green Approved by: re (scottl) END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-10-28T20:29:42.000000Z K 7 svn:log V 109 First fill in structure with valid values, and only then attach it to the global list. Reviewed by: rwatson END K 10 svn:author V 2 ru K 8 svn:date V 27 2005-10-28T21:25:28.000000Z K 7 svn:log V 77 Add some significant Ukrainian dates. PR: 88076 Submitted by: Andriy Gapon END K 10 svn:author V 3 sos K 8 svn:date V 27 2005-10-28T21:41:42.000000Z K 7 svn:log V 123 Return the right format (LBA/MSF) in TOC entries. Null the TOC on media change even if no media present. Reported by: des END K 10 svn:author V 6 brooks K 8 svn:date V 27 2005-10-28T21:57:04.000000Z K 7 svn:log V 64 Don't crash when given an invalid nwkey string. PR: misc/88159 END K 10 svn:author V 6 avatar K 8 svn:date V 27 2005-10-28T22:45:52.000000Z K 7 svn:log V 326 MFC: (1.16) Fixing a name collision in module declaration by renaming "firewire" to "if_firewire." According to the PR originator and Cai's test, this bug prevents a RELENG_6 GENERIC kernel from loading if_fwip.ko and will panic the kernel if users compile "device fwip" into their GENERIC kernel. Approved by: re (scottl) END K 10 svn:author V 6 brooks K 8 svn:date V 27 2005-10-28T22:47:40.000000Z K 7 svn:log V 237 s/insure/ensure/ in previous commit. My dictionary and m-w.com say they are synonymous, but ensure seems slightly closer and does not have the connotation of buying insurance. Reported by: Jason McIntyre END K 10 svn:author V 6 avatar K 8 svn:date V 27 2005-10-28T22:52:31.000000Z K 7 svn:log V 336 MFC: (1.16) Fixing a name collision in module declaration by renaming "firewire" to "if_firewire." According to the PR originator and Cai's testing result, this bug prevents a RELENG_6 GENERIC kernel from loading if_fwip.ko and will panic the kernel if users compile "device fwip" into their GENERIC kernel. Approved by: re (scottl) END K 10 svn:author V 6 brooks K 8 svn:date V 27 2005-10-28T22:57:42.000000Z K 7 svn:log V 187 MFC rev 1.6: Don't print anything if we can't do any localpkg shutdown (start already does this). Submitted by: Andre Albsmeier PR: conf/86606 END K 10 svn:author V 5 wpaul K 8 svn:date V 27 2005-10-29T02:18:27.000000Z K 7 svn:log V 239 Fix ndis_getstate_80211() so that it properly reports the authmode and channel to ifconfig. Also use the SSID and channel info from the association info that we already have instead of using ndis_get_info() to ask the driver for it again. END K 10 svn:author V 5 wpaul K 8 svn:date V 27 2005-10-29T02:20:44.000000Z K 7 svn:log V 126 MFC: fix channel and authmode reporting for ifconfig (note: authmode reporting is a no-op in 5.x and always shows up as OPEN) END K 10 svn:author V 6 mjacob K 8 svn:date V 27 2005-10-29T02:46:59.000000Z K 7 svn:log V 133 Add an ioctl framework for doing FC task management functions from a user space tool- useful for doing FC target mode certification. END