summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* MFV: r276862jkim2015-01-092-2/+2
| | | | Fix build.
* Merge OpenSSL 1.0.1k.jkim2015-01-08443-7955/+9107
|\
| * Import OpenSSL 1.0.1k.jkim2015-01-08109-3632/+1762
| |
* | Now compiler-rt has been updated in r276851, bring in the variousdim2015-01-088-0/+220
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | sanitizer libraries that already work on FreeBSD: * asan: Address Sanitizer * ubsan: Undefined Behavior Sanitizer * profile: Profile Guided Optimization support Please note that these libraries are *experimental* at this stage, so the main Makefile is not yet connected to the build. Since I didn't want to needlessly edit BSD.usr.dist, you will also have to create the install directory /usr/lib/clang/3.5.0/lib/freebsd manually for now.
* | Change the default method for device_quiesce() to return 0 instead ofjhb2015-01-081-1/+1
| | | | | | | | | | | | | | | | | | EOPNOTSUPP. The current behavior can mask real quiesce errors since devclass_quiesce_driver() stops iterating over drivers as soon as it gets an error (incluiding EOPNOTSUPP), but the caller it returns the error to explicitly ignores EOPNOTSUPP. Reviewed by: imp
* | Apparently more Makefiles use stuff from compiler-rt, so fix them updim2015-01-083-5/+5
| | | | | | | | too. (This did not show during a make universe, strangely.)
* | Follow up to r276851 by also committing a necessary update todim2015-01-081-1/+1
| | | | | | | | | | | | lib/libblocksruntime. Pointy hat to: dim
* | Update compiler-rt to trunk r224034. This brings a number of newdim2015-01-08651-1760/+97789
|\ \ | | | | | | | | | | | | builtins, and also the various sanitizers. Support for these will be added in a later commit.
| * | Import compiler-rt trunk r224034.dim2015-01-072209-110277/+143965
| | | | | | | | | | | | https://llvm.org/svn/llvm-project/compiler-rt/trunk@224034
| * | Import compiler-rt release_34 branch r197381.dim2014-11-06679-7815/+36733
| | | | | | | | | | | | https://llvm.org/svn/llvm-project/compiler-rt/branches/release_34@197381
| * | Import compiler-rt r182741.ed2013-05-27434-8908/+22298
| | |
| * | Import compiler-rt r172839.andrew2013-01-18437-6095/+25435
| | |
* | | Reject attempts to read the cpuset mask of a negative domain ID.jhb2015-01-081-1/+1
| | |
* | | Document CPU_WHICH_DOMAIN and bump Dd for cpuset.1.jhb2015-01-082-2/+3
| | | | | | | | | | | | Missed in: r276829
* | | Revert r274569. It seems to be causing a crash when merging CTF data formarkj2015-01-081-1/+1
| | | | | | | | | | | | | | | | | | recent i386 GENERIC kernels. Reported by: David Wolfskill
* | | Add forgotten definition for nd6_output_ifp().melifaro2015-01-081-0/+2
| | |
* | | Add infrastructure to build dtb files from dts files.imp2015-01-085-1/+103
| | |
* | | * Use newly-created nd6_grab_holdchain() function to retrieve llemelifaro2015-01-083-207/+101
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | hold mbuf chain instead of calling full-blown nd6_output_lle() for each packet. This simplifies both callers and nd6_output_lle() implementation. * Make nd6_output_lle() static and remove now-unused lle and chain arguments. * Rename nd6_output_flush() -> nd6_flush_holdchain() to be consistent. * Move all pre-send transmit hooks to newly-created nd6_output_ifp(). Now nd6_output(), nd6_output_lle() and nd6_flush_holdchain() are using it to send mbufs to if_output. * Remove SeND hook from nd6_na_input() because it was implemented incorrectly since the beginning (r211501): - it tagged initial input mbuf (m) instead of m_hold - tagging _all_ mbufs in holdchain seems to be wrong anyway.
* | | We don't want tests on these embedded boxes.imp2015-01-081-0/+1
| | |
* | | Close a race in the isp(4) driver that caused devices to disappearken2015-01-082-7/+44
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | and not automatically come back if they were gone for a short period of time. The isp(4) driver has a 30 second gone device timer that gets activated whenever a device goes away. If the device comes back before the timer expires, we don't send a notification to CAM that it has gone away. If, however, there is a command sent to the device while it is gone and before it comes back, the isp(4) driver sends the command back with CAM_SEL_TIMEOUT status. CAM responds to the CAM_SEL_TIMEOUT status by removing the device. In the case where a device comes back within the 30 second gone device timer window, though, we weren't telling CAM the device came back. So, fix this by tracking whether we have told CAM the device is gone, and if we have, send a rescan if it comes back within the 30 second window. ispvar.h: In the fcportdb_t structure, add a new bitfield, reported_gone. This gets set whenever we return a command with CAM_SEL_TIMEOUT status on a Fibre Channel device. isp_freebsd.c: In isp_done(), if we're sending CAM_SEL_TIMEOUT for for a command sent to a FC device, set the reported_gone bit. In isp_async(), in the ISPASYNC_DEV_STAYED case, rescan the device in question if it is mapped to a target ID and has been reported gone. In isp_make_here(), take a port database entry argument, and clear the reported_gone bit when we send a rescan to CAM. In isp_make_gone(), take a port database entry as an argument, and set the reported_gone bit when we send an async event telling CAM consumers that the device is gone. Sponsored by: Spectra Logic MFC after: 1 week
* | | Do not initialize pfi_unlnkdkifs_mtx and pf_frag_mtx.rodrigc2015-01-082-8/+0
| | | | | | | | | | | | | | | | | | They are already initialized by MTX_SYSINIT. Submitted by: Nikos Vassiliadis <nvass@gmx.com>
* | | Need to include opt_cpu.h to access CPU_DISABLE_SSE option. Thankfully, thisimp2015-01-081-0/+2
| | | | | | | | | | | | | | | | | | only broke i686 configs that disabled SSE. Submitted by: nyan@
* | | Fix Fibre Channel Command Reference Number handling in the isp(4) driver.ken2015-01-082-0/+34
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The Command Reference Number is used for precise delivery of commands, and is part of the FC-Tape functionality set. (This is only enabled for devices that support precise delivery of commands.) It is an 8-bit unsigned number that increments from 1 to 255. The commands sent by the initiator must be processed by the target in CRN order if the CRN is non-zero. There are certain scenarios where the Command Reference Number sequence needs to be reset. When the target is power cycled, for instance, the initiator needs to reset the CRN to 1. The initiator will know this because it will see a LIP (when directly connected) or get a logout/login event (when connected to a switch). The isp(4) driver was not resetting the CRN when a target went away and came back. When it saw the target again after a power cycle, it would continue the CRN sequence where it left off. The target would ignore the command because the CRN sequence is supposed to be reset to 1 after a power cycle or other similar event. The symptom that the user would see is that there would be lots of aborted INQUIRY commands after a tape library was power cycled, and the library would fail to probe. The INQUIRY commands were being ignored by the tape drive due to the CRN issue mentioned above. isp_freebsd.c: Add a new function, isp_fcp_reset_crn(). This will reset all of the CRNs for a given port, or the CRNs for all LUNs on a target. Reset the CRNs for all targets on a port when we get a LIP, loop reset, or loop down event. Reset the CRN for a particular target when it arrives, is changed or departs. This is less precise behavior than the clearing behavior specified in the FCP-4 spec (which says that it should be reset for PRLI, PRLO, PLOGI and LOGO), but this is the level of information we have here. If this is insufficient, then we will need to add more precise notification from the lower level isp(4) code. isp_freebsd.h: Add a prototype for isp_fcp_reset_crn(). Sponsored by: Spectra Logic MFC after: 1 week
* | | Partial revert of r276832:pfg2015-01-081-0/+2
| | | | | | | | | | | | | | | | | | Do not bump the warns level as it still breaks the gcc build on sparc64 Reported by: jenkins
* | | Improve camcontrol(8) handling of drive defect data.ken2015-01-084-193/+542
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This includes a new summary mode (-s) for camcontrol defects that quickly tells the user the most important thing: how many defects are in the requested list. The actual location of the defects is less important. Modern drives frequently have more than the 8191 defects that can be reported by the READ DEFECT DATA (10) command. If they don't have that many grown defects, they certainly have more than 8191 defects in the primary (i.e. factory) defect list. The READ DEFECT DATA (12) command allows for longer parameter lists, as well as indexing into the list of defects, and so allows reporting many more defects. This has been tested with HGST drives and Seagate drives, but does not fully work with Seagate drives. Once I have a Seagate spec I may be able to determine whether it is possible to make it work with Seagate drives. scsi_da.h: Add a definition for the new long block defect format. Add bit and mask definitions for the new extended physical sector and bytes from index defect formats. Add a prototype for the new scsi_read_defects() CDB building function. scsi_da.c: Add a new scsi_read_defects() CDB building function. camcontrol(8) was previously composing CDBs manually. This is long overdue. camcontrol.c: Revamp the camcontrol defects subcommand. We now go through multiple stages in trying to get defect data off the drive while avoiding various drive firmware quirks. We start off by requesting the defect header with the 10 byte command. If we're in summary mode (-s) and the drive reports fewer defects than can be represented in the 10 byte header, we're done. Otherwise, we know that we need to issue the 12 byte command if the drive reports the maximum number of defects. If we're in summary mode, we're done if we get a good response back when asking for the 12 byte header. If the user has asked for the full list, then we use the address descriptor index field in the 12 byte CDB to step through the list in 64K chunks. 64K is small enough to work with most any ancient or modern SCSI controller. Add support for printing the new long block defect format, as well as the extended physical sector and bytes from index formats. I don't have any drives that support the new formats. Add a hexadecimal output format that can be turned on with -X. Add a quiet mode (-q) that can be turned on with the summary mode (-s) to just print out a number. Revamp the error detection and recovery code for the defects command to work with HGST drives. Call the new scsi_read_defects() CDB building function instead of rolling the CDB ourselves. Pay attention to the residual from the defect list request when printing it out, so we don't run off the end of the list. Use the new scsi_nv library routines to convert from strings to numbers and back. camcontrol.8: Document the new defect formats (longblock, extbfi, extphys) and command line options (-q, -s, -S and -X) for the defects subcommand. Explain a little more about what drives generally do and don't support. Sponsored by: Spectra Logic MFC after: 1 week
* | | Update Features2 to display SDBG capability of processor. This issbruno2015-01-082-1/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | showing up on Haswell-class CPUs From the Intel SDM, "Table 3-20. Feature Information Returned in the ECX Register" 11 | SDBG | A value of 1 indicates the processor supports IA32_DEBUG_INTERFACE MSR for silicon debug. Submitted by: jiashiun@gmail.com Reviewed by: jhb neel MFC after: 2 weeks
* | | Make /net use -intr by default. Linux does that, and it seems a good idea.trasz2015-01-081-1/+1
| | | | | | | | | | | | | | | MFC after: 1 month Sponsored by: The FreeBSD Foundation
* | | sed: Address warnings with clang and gcc48.pfg2015-01-083-7/+5
| | | | | | | | | | | | MFC after: 2 weeks
* | | Fix a bug in the CAM SCSI probe code that caused changes in inquiryken2015-01-081-1/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | data to go undetected. The probe code does an MD5 checksum of the inquiry data (and page 0x80 serial number if available) before doing a reprobe of an existing device, and then compares a checksum after the probe to see whether the device has changed. This check was broken in January, 2000 by change 56146 when the extended inquiry probe code was added. In the extended inquiry probe case, it was calculating the checksum a second time. The second time it included the updated inquiry data from the short inquiry probe (first 36 bytes). So it wouldn't catch cases where the vendor, product, revision, etc. changed. This change will have the effect that when a device's inquiry data is updated and a rescan is issued, it will disappear and then reappear. This is the appropriate action, because if the inquiry data or serial number changes, it is either a different device or the device configuration may have changed significantly. (e.g. with updated firmware.) scsi_xpt.c: Don't calculate the initial MD5 checksum on standard inquiry data and the page 0x80 serial number if we have already calculated it. MFC after: 1 week Sponsored by: Spectra Logic
* | | Create a cpuset mask for each NUMA domain that is available in thejhb2015-01-086-19/+49
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | kernel via the global cpuset_domain[] array. To export these to userland, add a CPU_WHICH_DOMAIN level that can be used to fetch the mask for a specific domain. Add a -d flag to cpuset(1) that can be used to fetch the mask for a given domain. Differential Revision: https://reviews.freebsd.org/D1232 Submitted by: jeff (kernel bits) Reviewed by: adrian, jeff
* | | Switch to x11/gnome3 now that x11/gnome2 nogjb2015-01-081-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | longer exists in the ports tree. MFC after: 3 days X-MFC-With: r276820, r276822, r276827 Sponsored by: The FreeBSD Foundation
* | | Ensure the ports directories exist for the list ofgjb2015-01-081-1/+20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | packages intended to be included on the DVD, and remove any nonexistent ports from the final list. Print the list of missing paths, and ensure that DVD_PACKAGES is non-zero length (which should never happen). MFC after: 3 days X-MFC-With: r276820, r276822 Sponsored by: The FreeBSD Foundation
* | | Drop 'new' from the description of NFSCL to match the recent change injhb2015-01-083-3/+3
| | | | | | | | | | | | HEAD kernel confs.
* | | Allow a block size of zero to mean 512 bytes, which is the most commonhselasky2015-01-081-1/+3
| | | | | | | | | | | | | | | | | | | | | block size for USB disks. This fixes support for "Action Cam SJ4000". Reviewed by: mav @ MFC after: 1 week
* | | Use the M_SIZE() macro when possible.hselasky2015-01-082-4/+2
| | | | | | | | | | | | | | | MFC after: 3 days Suggested by: rwatson@
* | | Add makefile for the "osmtest" utility. While at it:hselasky2015-01-086-6/+49
| | | | | | | | | | | | | | | | | | | | | | | | | | | - Fix depend target by removing a space after an "-I" inclusion option. - Fix some minor compile issues in the "osmtest" utility. MFC after: 3 days PR: 196580 Sponsored by: Mellanox Technologies
* | | Ensure /var/run/ld-elf.so.hints exists before buildinggjb2015-01-081-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | pkg(8). MFC after: 3 days X-MFC-With: r276820 Sponsored by: The FreeBSD Foundation
* | | nm: Accept long option --extern-only for -gemaste2015-01-082-3/+8
| | | | | | | | | | | | Submitted by: jkim
* | | Print informational output when NOPORTS is set,gjb2015-01-081-0/+3
| | | | | | | | | | | | | | | | | | | | | | | | which would otherwise cause pkg-stage.sh to silently exit. MFC after: 3 days Sponsored by: The FreeBSD Foundation
* | | Fix WITH_LLDB buildemaste2015-01-081-1/+1
| | | | | | | | | | | | | | | | | | A build with WITH_LLDB and not WITH_CLANG_EXTRAS failed after r276783. Reported by: rodrigc
* | | Replace hand-crafted versions of M_SIZE() and M_START() in uipc_mbuf.crwatson2015-01-081-7/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | with calls to the centralised macros, reducing direct use of MLEN and MHLEN. Differential Revision: https://reviews.freebsd.org/D1444 Reviewed by: bz Sponsored by: EMC / Isilon Storage Division
* | | Fix for compilation issue. Don't use the "abs()" function for unsignedhselasky2015-01-081-1/+5
| | | | | | | | | | | | | | | | | | | | | | | | computations. MFC after: 3 days PR: 196597 Sponsored by: Mellanox Technologies
* | | Correct comparison of IPv6 wildcard address.ume2015-01-081-1/+1
| | | | | | | | | | | | MFC after: 3 days
* | | Add myself (arybchik) to committers-src.dotarybchik2015-01-081-0/+2
| | | | | | | | | | | | Approved by: gnn (mentor)
* | | Move the inclusion of cpu-v6.h inside the #ifdef _KERNEL block, so thatian2015-01-081-4/+3
| | | | | | | | | | | | userland programs (which probably don't actually need machine/cpu.h) compile.
* | | Remove unnecessary .include of bsd.own.mkngie2015-01-081-2/+0
| | | | | | | | | | | | | | | MFC after: 3 days Sponsored by: EMC / Isilon Storage Division
* | | Build contrib/ofed/usr.{bin,lib} in parallelngie2015-01-082-4/+4
| | | | | | | | | | | | | | | MFC after: 1 week Sponsored by: EMC / Isilon Storage Division
* | | Fix 'make depend' before infiniband headers have been installed to build hostngie2015-01-081-1/+1
| | | | | | | | | | | | | | | | | | | | | by removing space between -I and the header directory MFC after: 1 week Sponsored by: EMC / Isilon Storage Division
* | | Add accessors for the ARM CP15 performance monitor registers. Also ensureian2015-01-083-35/+73
| | | | | | | | | | | | | | | | | | that some #ifdef SMP code is also conditional on __ARM_ARCH >= 7; we don't support SMP on armv6, but some drivers and modules are compiled with it forced on via the compiler command line.
* | | Clarify text to be consistent with nanosleep(2),rodrigc2015-01-081-4/+4
| | | | | | | | | | | | | | | | | | since sleep(3) is implemented in terms of nanosleep(2). This is similar to the sleep(3) man page for Darwin.
OpenPOWER on IntegriCloud