aboutsummaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
...
* - Add the character "E" to the understood lock types. This meansJeff Roberson2005-04-111-0/+2
| | | | | | | | | the VOP requires an exclusive lock. Sponsored by: Isilon Systems, Inc. Notes: svn path=/head/; revision=144907
* Use comments after .endif to please make(1) with latest changes soMaxime Henrion2005-04-111-2/+2
| | | | | | | | that it's possible to build a kernel without getting flooded with thousands of warnings. Notes: svn path=/head/; revision=144906
* - Clear VI_OWEINACT before calling vget() with no lock type. We knowJeff Roberson2005-04-111-0/+7
| | | | | | | | the node is actually already locked, and VOP_INACTIVE is not desirable in this case. Notes: svn path=/head/; revision=144904
* - Honor the flags argument passed to null_root(). The filesystem belowJeff Roberson2005-04-111-1/+1
| | | | | | | us will decide whether or not to grab a real shared lock. Notes: svn path=/head/; revision=144903
* Replace buggy for-loops to skip certain character with strspn(). If *fmt wasStefan Farfeleder2005-04-111-3/+3
| | | | | | | | | | '\0' (eg in the invocation 'printf %'), the for-loop would miss the terminating null character. MFC after: 1 week Notes: svn path=/head/; revision=144902
* Remove goto.Gleb Smirnoff2005-04-111-20/+16
| | | | Notes: svn path=/head/; revision=144901
* - Change the VOP_LOCK UPGRADE in vput() to do a LK_NOWAIT to avoid aJeff Roberson2005-04-111-39/+43
| | | | | | | | | | | | | | | potential lock order reversal. Also, don't unlock the vnode if this fails, lockmgr has already unlocked it for us. - Restructure vget() now that vn_lock() does all of VI_DOOMED checking for us and also handles the case where there is no real lock type. - If VI_OWEINACT is set, we need to upgrade the lock request to EXCLUSIVE so that we can call inactive. It's not legal to vget a vnode that hasn't had INACTIVE called yet. Sponsored by: Isilon Systems, Inc. Notes: svn path=/head/; revision=144900
* - Assert that we're no longer doing recursive vn_locks in inactive/reclaimJeff Roberson2005-04-111-3/+12
| | | | | | | | | | | | | | | | as I'd like to get rid of the vxthread. - Handle lock requests which don't actually want a lock as this is a much more convenient place to handle this condition than in vget(). These requests simply want to know that VI_DOOMED isn't set. - Correct a test at the end of vn_lock, if error !=0 should be if error == 0, this has been broken since I comitted the VI_DOOMED changes, but no one ran into it because vget() duplicated this functionality. Sponsored by: Isilon Systems, Inc. Notes: svn path=/head/; revision=144899
* - vput(tvp) before vrele(tdvp) in kern_rename() to avoid lock order issues.Jeff Roberson2005-04-112-4/+4
| | | | Notes: svn path=/head/; revision=144898
* Parse_Init() is gone so remove comments that mention it.Hartmut Brandt2005-04-111-11/+0
| | | | Notes: svn path=/head/; revision=144897
* Remove an unneccesary forward declaration for MainParseArgs().Hartmut Brandt2005-04-111-20/+34
| | | | | | | | | | | | | getopt() may be called several times - make sure to set optreset to reset it. Cleanup handling of non-option arguments. Remove some misleading comments. Patch: 7.171 Submitted by: Max Okumoto <okumoto@ucsd.edu> Notes: svn path=/head/; revision=144896
* Remove the call to Parse_Init() - there is no need to initializeHartmut Brandt2005-04-111-2/+0
| | | | | | | | a static variable to NULL. Forgot this in the previous commit to parse.[ch]. Notes: svn path=/head/; revision=144895
* Rework the directive parsing code. Instead of using a lot of strcmp()sHartmut Brandt2005-04-116-599/+673
| | | | | | | | | | | | | | | on every line that starts with a dot use a minimal perfect hash function and a single strcmp() on the first word after the dot to find out whether it is really a directive call and, if yes, which one. Then directly dispatch to a handler function for that directive (or fall through to the dependency handling code). This makes the directive parse a little bit more strict about the syntax: the directive word must be followed by a character that is not alphanumerical and not an underline (making .undefFOO illegal); .endif and .else can only be followed by comments. Notes: svn path=/head/; revision=144894
* Make sure the only thing that follows .endif or .else is a comment.Hartmut Brandt2005-04-118-14/+14
| | | | Notes: svn path=/head/; revision=144893
* Increase PTHREAD_STACK_MIN to MINSIGSTKSZ because thread libraries now useDavid Xu2005-04-111-1/+1
| | | | | | | | | makecontext which enforces a minimum stack size to be MINSIGSTKSZ. Bug report: Bill Middleton <flashdict at gmail dot com>, BSD-sharp project. Notes: svn path=/head/; revision=144892
* The alternative suggested for /entropy as a shutdownDoug Barton2005-04-112-6/+5
| | | | | | | | | | | | | | | | | | save file was /var/db/entropy, which also happens to be the directory where the individual entropy files created by /usr/libexec/save-entropy are stored. Change the suggestion to be /var/db/entropy-file instead. In an error condition where the shutdown file is not created, the error message accessed a variable that doesn't exist. PR: conf/75722 Submitted by: Nicolas Rachinsky <list@rachinsky.de> Notes: svn path=/head/; revision=144891
* It isn't a whinable offence to want memory when the bar says ioport.Warner Losh2005-04-111-6/+8
| | | | | | | | | | Put that behind bootverbose to make the ata driver less chatty on advanced hardware. Requested by: sos Notes: svn path=/head/; revision=144890
* The script mistakenly ignores the entropy_save_sz variable fromDoug Barton2005-04-111-2/+2
| | | | | | | | | | | | | | | | | | rc.conf[.local]. Fix this, and leave the default as 2048. Update the copyright year to include the present. Update the assignment of the copyright to be me personally, instead of "The FreeBSD Project" which is not a legal entity, and therefore not a proper assignee. My intention remains the same however, that this code continue to be BSD licensed, and freely available to anyone that wants it under those terms. PR: conf/75722 Submitted by: Nicolas Rachinsky <list@rachinsky.de> Notes: svn path=/head/; revision=144889
* Create new i386 windows/bsd thunking layer, similar to the amd64 thunkingBill Paul2005-04-1113-965/+1727
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | layer, but with a twist. The twist has to do with the fact that Microsoft supports structured exception handling in kernel mode. On the i386 arch, exception handling is implemented by hanging an exception registration list off the Thread Environment Block (TEB), and the TEB is accessed via the %fs register. The problem is, we use %fs as a pointer to the pcpu stucture, which means any driver that tries to write through %fs:0 will overwrite the curthread pointer and make a serious mess of things. To get around this, Project Evil now creates a special entry in the GDT on each processor. When we call into Windows code, a context switch routine will fix up %fs so it points to our new descriptor, which in turn points to a fake TEB. When the Windows code returns, or calls out to an external routine, we swap %fs back again. Currently, Project Evil makes use of GDT slot 7, which is all 0s by default. I fully expect someone to jump up and say I can't do that, but I couldn't find any code that makes use of this entry anywhere. Sadly, this was the only method I could come up with that worked on both UP and SMP. (Modifying the LDT works on UP, but becomes incredibly complicated on SMP.) If necessary, the context switching stuff can be yanked out while preserving the convention calling wrappers. (Fortunately, it looks like Microsoft uses some special epilog/prolog code on amd64 to implement exception handling, so the same nastiness won't be necessary on that arch.) The advantages are: - Any driver that uses %fs as though it were a TEB pointer won't clobber pcpu. - All the __stdcall/__fastcall/__regparm stuff that's specific to gcc goes away. Also, while I'm here, switch NdisGetSystemUpTime() back to using nanouptime() again. It turns out nanouptime() is way more accurate than just using ticks(). On slower machines, the Atheros drivers I tested seem to take a long time to associate due to the loss in accuracy. Notes: svn path=/head/; revision=144888
* Go ahead and try to allocate PCI_BAR(5) for ata devices.Warner Losh2005-04-101-0/+1
| | | | Notes: svn path=/head/; revision=144887
* Use the ata_suspend/resume functions instaead of the bus_generic ones.Søren Schmidt2005-04-101-2/+2
| | | | | | | | | This should unbreak suspend/resume. Contributed by: Wiktor Niesiobedzki Notes: svn path=/head/; revision=144886
* Fix FreeBSD native SPAN arrays.Søren Schmidt2005-04-101-8/+8
| | | | | | | Prodded by: Ian Dowse Notes: svn path=/head/; revision=144885
* Align the entry point to assembly language functions to a 16-byte boundary.Alan Cox2005-04-101-2/+1
| | | | | | | | (The Opteron's instruction fetcher reads instructions from the L1 cache in 16-byte, aligned packets.) Notes: svn path=/head/; revision=144884
* Drop the default running value to 60% from 80%. Testing shows that weNate Lawson2005-04-102-3/+3
| | | | | | | | cycle too much, wasting power if we trigger a jump to full speed too easily. CPU-bound tasks quickly push idle below this level. Notes: svn path=/head/; revision=144883
* Set CPU speed to 100% in acpi_throttle attach. This is needed for someNate Lawson2005-04-102-3/+12
| | | | | | | | | | | | systems that boot with this value at the lowest setting. Change the default boot config back to "leave frequency as BIOS set it". Also, fix buglet where acpi_throttle wouldn't be used if p4tcc was present but disabled by the user. MFC after: 1 week Notes: svn path=/head/; revision=144882
* Properly terminate the table generated from ACPI info. The cpufreqNate Lawson2005-04-101-6/+6
| | | | | | | | settings are length-counted while the EST table is null-terminated. This fixes extra garbage states being reported with ACPI probing. Notes: svn path=/head/; revision=144881
* Update for verbose debugging support and add Bruno Ducrot author note.Nate Lawson2005-04-101-6/+10
| | | | Notes: svn path=/head/; revision=144879
* Advertise that we can handle unified SMP control of processor powerNate Lawson2005-04-101-2/+4
| | | | | | | states, idling, etc. This has been supported since the cpufreq import. Notes: svn path=/head/; revision=144878
* Advertise p4tcc via acpi_get_features() _PDC support.Nate Lawson2005-04-101-0/+18
| | | | Notes: svn path=/head/; revision=144877
* Add debugging prints to all the methods in case there are problems withNate Lawson2005-04-101-7/+70
| | | | | | | | managing levels. This can be enabled with the debug.cpufreq.verbose tunable and sysctl. Notes: svn path=/head/; revision=144876
* Fix support for _PDC by using the proper version/length format for theNate Lawson2005-04-102-3/+24
| | | | | | | | buffer. Also, reference the Intel document where the _PDC values were found. This now supports ACPI-assisted SpeedStep on my borrowed T42. Notes: svn path=/head/; revision=144875
* Add myself to the calendar.Remko Lodder2005-04-101-0/+1
| | | | | | | | MFC after: 1 week Approved by: simon (mentor) Notes: svn path=/head/; revision=144874
* Add support for recording the total energy used (in joules) when inNate Lawson2005-04-101-6/+47
| | | | | | | | | verbose mode. This is useful for profiling new adaptive algorithms in performance (via time(1)) and total energy consumed for a given workload. Notes: svn path=/head/; revision=144873
* Add a machine-specific, optimized implementation of strcat.Alan Cox2005-04-102-1/+166
| | | | | | | | | PR: 73111 Submitted by: Ville-Pertti Keinonen <will@iki.fi> (taken from NetBSD) MFC after: 3 weeks Notes: svn path=/head/; revision=144872
* Eliminate a conditional branch and as a side-effect eliminate a branch toAlan Cox2005-04-102-10/+6
| | | | | | | | | | | a return instruction. (The latter is discouraged by the Opteron optimization manual because it disables branch prediction for the return instruction.) Reviewed by: bde Notes: svn path=/head/; revision=144868
* Update copyright: parts of the netipx implementation are covered by aRobert Watson2005-04-101-1/+1
| | | | | | | | | 2005 copyright. MFC after: 3 days Notes: svn path=/head/; revision=144867
* Compare (mbuf *) with NULL, not 0.Robert Watson2005-04-101-1/+1
| | | | | | | MFC after: 3 days Notes: svn path=/head/; revision=144866
* Fix the output ports on the AD1988 codec. From the submitter:Scott Long2005-04-101-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | Affects to people WITH an AD1888 codec, the system will output to the port labeled "speaker" instead of microphone. System will work the same in multiple operating systems. If people are currently using their systems with this codec they will need to swap their output ports. I have _not_ checked audio input or line input (basically, I have checked nothing other than line-out). I believe this is an appropriate change, it makes us consistent with documentation, and other operating systems. Furthermore, this feature (playing) is the vast majority of sound activities, so if this makes is right for playback and wrong for recording... playback is more important, and we can fix recoding in the future without worries of screwing people again in the future (since we'll be "right" on the playback). Submitted by: David Cross Notes: svn path=/head/; revision=144865
* o Add DragonFly 1.2.0.Maxim Konovalov2005-04-101-3/+4
| | | | | | | o s/DFD/DFB/g as DragonFly people did. Notes: svn path=/head/; revision=144864
* Remove duplicated "bytes".Xin LI2005-04-101-1/+1
| | | | | | | | Submitted by: Wojciech A. Koszek [dunstan freebsd czest pl] PR: 79747 Notes: svn path=/head/; revision=144863
* s/self-distruction/self-destructionJens Schweikhardt2005-04-101-2/+2
| | | | Notes: svn path=/head/; revision=144862
* Refine the SATA PHY code so the time consumed during probe is reduced.Søren Schmidt2005-04-103-10/+15
| | | | | | | Add the older nVidia nForce2 SATA chip to supported HW. Notes: svn path=/head/; revision=144861
* - Tighten up the Timestamp checks to prevent a spoofed segment fromPaul Saab2005-04-103-6/+47
| | | | | | | | | | | | | setting ts_recent to an arbitrary value, stopping further communication between the two hosts. - If the Echoed Timestamp is greater than the current time, fall back to the non RFC 1323 RTT calculation. Submitted by: Raja Mukerji (raja at moselle dot com) Reviewed by: Noritoshi Demizu, Mohan Srinivasan Notes: svn path=/head/; revision=144858
* - If the reassembly queue limit was reached or if we couldn't allocatePaul Saab2005-04-104-2/+8
| | | | | | | | | | | | | | a reassembly queue state structure, don't update (receiver) sack report. - Similarly, if tcp_drain() is called, freeing up all items on the reassembly queue, clean the sack report. Found, Submitted by: Noritoshi Demizu <demizu at dd dot iij4u dot or dot jp> Reviewed by: Mohan Srinivasan (mohans at yahoo-inc dot com), Raja Mukerji (raja at moselle dot com). Notes: svn path=/head/; revision=144857
* When the rightmost SACK block expands, rcv_lastsack should be updated.Paul Saab2005-04-101-0/+2
| | | | | | | | | | | (Fix for kern/78226). Submitted by : Noritoshi Demizu <demizu at dd dot iij4u dot or dot jp> Reviewed by : Mohan Srinivasan (mohans at yahoo-inc dot com), Raja Mukerji (raja at moselle dot com). Notes: svn path=/head/; revision=144856
* Remove some unused sack fields.Paul Saab2005-04-101-3/+0
| | | | | | | Submitted by : Noritoshi Demizu, Mohan Srinivasan. Notes: svn path=/head/; revision=144855
* Add a machine-specific, optimized implementation of strcpy.Alan Cox2005-04-102-1/+113
| | | | | | | | | PR: 73111 Submitted by: Ville-Pertti Keinonen <will@iki.fi> (taken from NetBSD) MFC after: 3 weeks Notes: svn path=/head/; revision=144854
* Oops, correct typo.Yoshihiro Takahashi2005-04-107-7/+7
| | | | Notes: svn path=/head/; revision=144853
* Initialize vp before using it. Failing to do this can cause instantXin LI2005-04-101-0/+1
| | | | | | | | | panic when trying to access a file on mounted smbfs. Submitted by: takawata at jp freebsd org Notes: svn path=/head/; revision=144852
* Suspend all other threads in the process while generating a core dump.David Schultz2005-04-101-1/+11
| | | | | | | | | | | | | | | The main reason for doing this is that the ELF dump handler expects the thread list to be fixed while the dump header is generated, so an upcall that occurs at the wrong time can lead to buffer overruns and other Bad Things. Another solution would be to grab sched_lock in the ELF dump handler, but we might as well single-thread, since the process is about to die. Furthermore, I think this should ensure that the register sets in the core file are sequentially consistent. Notes: svn path=/head/; revision=144851
generated by cgit v1.2.3 (git 2.25.1) at 2024-10-13 02:30:18 +0000