summaryrefslogtreecommitdiffstats
path: root/sys
Commit message (Collapse)AuthorAgeFilesLines
...
* Added comment why this workaround is required.ume2003-01-221-1/+7
| | | | | Suggested by: sam MFC after: 1 week
* Rename cpuid_cpuinfo to cpu_procinfo. bde requested that I rename thisjhb2003-01-2215-19/+19
| | | | | | | variable to something in the cpu_* namespace since that's what all the other cpuid variables were named and cpu_procinfo is what I came up with. Requested by: bde
* Bah, add in a missing space char I noticed when MFC'ing this.jhb2003-01-222-2/+2
|
* Correctly identify the 3C920B chipset; although it may ship with some Nforce2silby2003-01-222-5/+5
| | | | | | boards, it's definitely not an Nvidia chip. Info from: Nvidia's Linux Network driver & pciids.sourceforge.net
* Make the arguments to d_close() match the arguments to d_open(), sincejoerg2003-01-221-1/+1
| | | | | | | | | otherwise the close operation is void, and the underlying device remains marked as being busy. Thanks to phk for finding this. MFC after: 3 days
* getpeername() returns with no error but didn't fill struct sockaddrume2003-01-221-0/+2
| | | | | | | | | | correctly against PF_LOCAL. It seems that the test always fails then sockaddr was not filled. So, I added else clause for workaround. I doubt if it is right fix. However, it is better than nothing. I found that NetBSD has same potential problem. But, fortunately, NetBSD has equivalent else clause. MFC after: 1 week
* Add yet another Sony Vaio PS/2 mouse ID. This one is found on a VX88.hsu2003-01-222-4/+6
| | | | Submitted by: marcel
* Remove BAT invalidation. This is done later in the boot sequence,grehan2003-01-222-34/+0
| | | | | | | so isn't required here, and seems to cause problems when booting from disk. Approved by: benno
* Convert remaining .s files to .Sgrehan2003-01-221-3/+3
| | | | Approved by: benno
* - remove dead conditional CFLAG settinggrehan2003-01-221-8/+3
| | | | | | - switch locore .s -> .S Approved by: benno
* There's absolutely no need for a struct-within-a-struct, so move thedes2003-01-212-28/+24
| | | | counters out of the inner struct and remove it.
* Add missing SMP file locks around read-modify-write operations onhsu2003-01-211-1/+5
| | | | | | the flag field. Reviewed by: rwatson
* Remove OBE prototype for iszerodev() - it was replaced by thepeter2003-01-212-2/+0
| | | | D_MMAP_ANON device switch flag.
* Fixes for a number of problems in the IOMMU code:tmm2003-01-212-16/+23
| | | | | | | | | | | | | | | | | | | | | | | | | | 1.) Fix an off-by-one in the DVMA space handling, which would make it possible to allocate one page beyond the end of the DVMA area. This page was aliased to the first page. Apparently, this bug was responsible for the trashed nvram/eeprom some people were reporting, in conjunction with a number of unfortunate coincidences. 2.) Fix broken boundary and and lowaddr calculations. 3.) Fix a memory leak on an error path. 4.) Update a outdated comment to reflect the introduction of IOMMU_MAX_PRE, make the usage of IOMMU_MAX_PRE more consistent and KASSERT that the preallocation size is not 0. 5.) Fix a case where an error return was lost. 6.) When signalling an error to the caller by invoking the callback, do not use a segment pointer of NULL for compatability with existing drivers. Also, increase the maximum segment number to 64; it is rather arbitrary, with the exception of the of the stack space consumed by the segment array. Special thanks go to Harti Brandt <brandt@fokus.fraunhofer.de> for spotting 4 and 5, and testing many iterations of patches. Pointy hats to: tmm
* Free resources when failing to set up the interrupt.tmm2003-01-211-0/+1
|
* bus_dmamap_sync() overhaul:tmm2003-01-211-21/+24
| | | | | | | | - Remove NetBSD-style or-ed together BUS_DMASYNC operations, in some cases relaxing the (intended) syncing operation a bit. - Add syncs before reading the descriptor rings. - Try to combine syncs where possible to avoid overhead. - Sync all maps before unloading them.
* bus_dmamap_sync() overhaul:tmm2003-01-212-22/+18
| | | | | | | | | | | - Remove NetBSD-style or-ed together BUS_DMASYNC operations, in some cases relaxing the (intended) syncing operation a bit. - Stop pretending that that we can sync part of a dmamap: replace the GEM_CDTXSYNC and GEM_CDRXSYNC macros with GEM_CDSYNC to sync the complete control map, and combine syncs wherever possible to avoid the overhead. - Sync all maps before unloading them. - Remove a few syncs which should be unnecessary.
* Fix iommu_dvmamap_sync(): it was still operating as if the BUS_DMASYNC_*tmm2003-01-211-5/+13
| | | | | | constants where flag bits (as in NetBSD), although they are consecutively numbered in FreeBSD. This would cause unnecessary flushing in the BUS_DMASYNC_POSTWRITE case, but was otherwise mostly harmless.
* Correct an off-by-one in the boundary check. Otherwise, resourcetmm2003-01-211-1/+1
| | | | | allocations would fail if the desired allocation size was equal to the boundary.
* Define misc structs outside of struct firewire_comm.simokawa2003-01-211-77/+85
|
* Ignore events of unused IR DMA.simokawa2003-01-211-3/+11
|
* Remove NEXTDOSPART.nyan2003-01-211-1/+0
|
* Use NDOSPART instead of NEXTDOSPART.nyan2003-01-211-2/+2
|
* MFi386: revision 1.63.nyan2003-01-211-10/+9
|
* Needs wlan for wi.imp2003-01-211-0/+1
|
* #ifdef NO_GEOM some bits here too.phk2003-01-211-0/+2
|
* Regenjoe2003-01-212-2/+9
|
* Add support for Epson 1660 scanner.joe2003-01-212-0/+2
| | | | | Submitted by: Bruno Schwander <bruno@tinkerbox.org> MFC after: 3 days
* use 'void *' instead of 'caddr_t' for useracc, kernacc, vslock and vsunlock.alfred2003-01-212-8/+8
|
* #ifdef NO_GEOM all of this file.phk2003-01-211-0/+2
|
* Bandaid to make the kernel compile until the scsi-crew can find out whatphk2003-01-211-1/+2
| | | | is happening.
* Add yet another Sony Vaio PS/2 mouse ID. This one is found in a R505GL.marcel2003-01-212-0/+2
| | | | | The ACPI data for the mouse includes the compatibility ID, but we apparently don't make use of it.
* Add a missing '{'phk2003-01-211-1/+1
|
* Remove M_TRYWAIT/M_WAITOK/M_WAIT. Callers should use 0.alfred2003-01-21591-2122/+2123
| | | | Merge M_NOWAIT/M_DONTWAIT into a single flag M_NOWAIT.
* Remove the fake disklabel, no longer needed.sos2003-01-212-47/+7
| | | | | | Fix potential malloc panic. Prodded by: phk
* preserve the order of tags copied by m_tag_copy_chainsam2003-01-211-3/+2
| | | | Obtained from: OpenBSD
* Run kldxref on sparc64, it works now.jake2003-01-212-2/+2
|
* Fix module dependency (pre)loading on sparc64 by relocating the variablesjake2003-01-211-6/+57
| | | | | | | | read from the raw kld files. Submitted by: Hartmut Brandt <brandt@fokus.gmd.de> PR: 46870 Tested on: alpha (obrien), i386, sparc64
* - Add a VM_WAIT in the appropriate cases where vm_page_alloc() fails and flagsjeff2003-01-213-31/+50
| | | | | | | | | | indicate that uma_small_alloc should not. This code should be refactored so that there is not so much cross arch duplication. Reviewed by: jake Spotted by: tmm Tested on: alpha, sparc64 Pointy hat to: jeff and everyone who cut and pasted the bad code. :-)
* Rewrite the SMP filedesc locking in knote_attach() in order tohsu2003-01-211-23/+11
| | | | | | | | | | | | | 1. eliminate unnecessary loop which frees and re-allocates the just allocated array 2. eliminate the newsize recomputation 3. eliminate unnecessary unlock and relock around free 4. correctly match the free with the malloc into M_KQUEUE instead of M_TEMP 5. eliminate conditional assignment of oldlist, which is equivalent to a simple assignment 6. eliminate the oldlist temporary variable completely Reviewed by: jhb
* GC an unused reference to vop_refreshlabel_desc; reference torwatson2003-01-211-3/+0
| | | | | | | opt_mac.h was removed previously so it was never compiled in. Obtained from: TrustedBSD Project Sponsored by: DARPA, Network Associates Laboratories
* Perform VOP_GETATTR() before mac_check_vnode_exec() so thatrwatson2003-01-211-5/+5
| | | | | | | the cached attributes are available to MAC modules. Submitted by: mike halderman <mrh@nosc.mil> Obtained from: TrustedBSD Project
* Resolve relative relocations in klds before trying to parse the module'sjake2003-01-2110-96/+232
| | | | | | | | | | | | metadata. This fixes module dependency resolution by the kernel linker on sparc64, where the relocations for the metadata are different than on other architectures; the relative offset is in the addend of an Elf_Rela record instead of the original value of the location being patched. Also fix printf formats in debug code. Submitted by: Hartmut Brandt <brandt@fokus.gmd.de> PR: 46732 Tested on: alpha (obrien), i386, sparc64
* Fix a missed goal.period -> goal.offset change. Ingibbs2003-01-201-2/+2
| | | | | this case, the bug resulted in comparing a period against an offset.
* include cdefs.h so as to unbreak the libc buildgallatin2003-01-201-0/+2
|
* move sysctl's under hw.hifn; change max batching default to 1 (no batching)sam2003-01-201-7/+9
| | | | based on what I learned from the Broadcom h/w
* move sysctl's under hw.ubsecsam2003-01-201-9/+10
|
* correct default setting of hw.wi.txerate; it must be -1, not zero,sam2003-01-201-2/+2
| | | | to silence all msgs (like the old driver)
* correct sysctl names and move them to hw.wi: hw.wi.txerate controls the ratesam2003-01-201-4/+6
| | | | | | | | at which tx errors are printed (default to 0); hw.wi.debug control the debug msgs and is only present when WI_DEBUG is defined at compile time (the default for the moment) Requested by: imp
* aic7xxx.reg:gibbs2003-01-206-138/+251
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add a constant for the controller's stack size and the maximum scsi offset. aic7xxx.seq: Style nit. The source is implied to be the destination unless overridden in an "and" instruction. Update target mode code for changes in identify seen sequencer flags. aic7xxx_pci.c: Ensure that the PCIERRGENDIS bit is set in the PCIERRGEN config space register. Perhaps this is a reason for the spurios parity errors reported on U160 controllers. Honor the AHC_NO_BIOS_INIT flag. Allow PCI interrupt reporting to be disabled, by clearing the PERRRESEN bit in the command register. This option is now enabled via a new softc flag: AHC_DISABLE_PCI_PERR. Disable SERR and pause the controller prior to performing our mmapped I/O test. This should handle the case of controllers that do not "auto-access pause". For legacy controllers, use SCB ram instead of scratch ram since the latter may contain settings left over from the BIOS that we will use if an seeprom is not found. Make use of new ahc_inl/outl() inlines. aic7xxx.h: Reformat a few comments to follow driver style. Add a controller flags that indicate that a controller has not been initialized by the BIOS and whether to disable PCI parity errors.. Remove stack probing softc members. Add a few more syncrate constants that are useful in speed fallback calculations. Add the SHOW_MASKED_ERRORS debug flag. aic7xxx.h: aic7xxx.c: Implement the SCB_SILENT flag. This is useful for hushing up the driver during DV or other operations that we expect to cause transmission errors. The messages will still print if the SHOW_MASKED_ERRORS debug option is enabled. aic7xxx_inline.h: Implement ahc_[in|out][w|l|q]. This removes the need for manual 'or and shift" type operations throughout the driver. aic7xxx.c: Move SELTO dignostic so that the SCB is still valid when we use it for printing path information. If we are narrow, limit syncrate to Ultra2. Don't clobber ppr_options when forcing a renegotiation. The current ppr_options may be referenced while queuing new commands. Don't set our width to unknown when forcing negotiation on narrow controllers. This will confuse the negotiation code into negotiating with a wide message on narrow controllers. Add an "asserting atn" diagnostic with controller/target information. Remove the probe_stack code. The stack is always 4 deep on legacy controllers, so probing is pointless. This also avoids an issue where probing the stack would upset the aic7770. In ahc_reset(), record whether or not we found the controller in a reset state. If the controller was already reset, assume that no BIOS has initialized the controller and ignore left over scratch ram settings. Fix an ifdef bug that caused sequencer debugging to be enabled always. Clear the ultraenb flag in our tstate during startup. The ultraenbled'ness of a device is recorded in the user transfer settings. tstate->ultraenb bitmask indicates which devices we have negotiated an ultra speed with. Just after initialization, we are async. Setting the ultraenb flag while async seems to be harmless, but it was confusing to see the ULTRAENB flag set in the SCB. Enhance residual diagnostic to indicate if the residual if for sense information or normal data transfers. Indicate the features, bugs, and flags set in the softc that are used to control firmware patch download when booting verbose. In ahc_dump_card_state() fix a logic reversal. The SCSIPHASE register only exists on U160 controllers. The SCSISIGI register exists on all controllers. Not the other way around. Also print out the ERROR register. Allow ahc_dump_card_state() to be called when the sequencer is not paused. Add dump card state markers as in the U320 driver.
OpenPOWER on IntegriCloud