dragonfly.git
7 years agokernel - Improve vm_page_register_action*() performance.
Matthew Dillon [Sun, 22 Jan 2017 23:28:05 +0000 (15:28 -0800)]
kernel - Improve vm_page_register_action*() performance.

* Improve the performance for vm_page_register_action() and related
  routines by splitting the global lock into per-hash-index locks.
  Also change from a token to lockmgr locks.

* Shift some code around in umtx_sleep() so the tsleep_interlock()
  occurs after the registration code to avoid interference with
  the new lockmgr() operations in the registration code.

7 years agovkernel - Fix vmspace_*() call bottleneck
Matthew Dillon [Sun, 22 Jan 2017 21:47:05 +0000 (13:47 -0800)]
vkernel - Fix vmspace_*() call bottleneck

* Remove a global token in most cases by caching ve's, and hold it shared
  for lookups if it cannot be found in the cache.

7 years agovkernel - Fix races in if_vke.c
Matthew Dillon [Sun, 22 Jan 2017 21:46:20 +0000 (13:46 -0800)]
vkernel - Fix races in if_vke.c

* Fix races in if_vke.c, several cpu_lfence() calls were in the wrong place.

7 years agoif_iwm - Add firmware API definitions for TX power commands.
Imre Vadász [Sun, 22 Jan 2017 21:38:31 +0000 (22:38 +0100)]
if_iwm - Add firmware API definitions for TX power commands.

* While there remove unused IWM_UCODE_TLV_CAPA_LMAC_UPLOAD definition,
  which isn't defined in iwlwifi.

Taken-From: Linux iwlwifi

7 years agokernel/if_wi: Unlock before returning.
Sascha Wildner [Sun, 22 Jan 2017 19:55:29 +0000 (20:55 +0100)]
kernel/if_wi: Unlock before returning.

7 years agovkernel - Fix pv_entry_count bug
Matthew Dillon [Sun, 22 Jan 2017 17:23:37 +0000 (09:23 -0800)]
vkernel - Fix pv_entry_count bug

* Now that we are not protected by a global vm_token, pv_entry_count
  must use atomic ops.

7 years agokernel - Fix NFS sillyrename code
Matthew Dillon [Sun, 22 Jan 2017 16:36:23 +0000 (08:36 -0800)]
kernel - Fix NFS sillyrename code

* Due to the delayed vnode reclamation changes done several years ago
  the NFS client code was not properly removing silly-renamed files
  on last-close (these are files which NFS renames to .nfs* when
  remove()'d while still open()).

* Fixed by flagging finalization on last-close if a sillyrename occurs.

7 years agosched: Add cpuset_t for FreeBSD compat
Sepherosa Ziehau [Sun, 22 Jan 2017 14:25:08 +0000 (22:25 +0800)]
sched: Add cpuset_t for FreeBSD compat

7 years agolibc: Add sched_getcpu().
Sepherosa Ziehau [Thu, 19 Jan 2017 14:03:01 +0000 (22:03 +0800)]
libc: Add sched_getcpu().

7 years agovkernel - Restabilize pmap code, redo kqueue, systimer, and console code
Matthew Dillon [Sun, 22 Jan 2017 07:43:07 +0000 (23:43 -0800)]
vkernel - Restabilize pmap code, redo kqueue, systimer, and console code

* Remove vm_token and add necessary vm_page spin locks to the vkernel's
  pmap code, improving its stability.

* Separate the systimer interrupt and console tty support from the
  kqueue subsystem.

  Uses SIGURG for systimer
  Uses SIGIO for kqueue
  Uses SIGALRM for cothread signalling

* The vkernel systimer code now uses a dedicated cothread for timing.  The
  cothread is a bit of a hack at the moment but is a more direct way of
  handling systimers.

* Attempt to fix user%/sys%/intr%/idle% in the systat -vm and
  systat -pv output.  Still isn't perfect, but it is now more
  accurate.

7 years agovkernel.7 - Update manual page
Matthew Dillon [Fri, 20 Jan 2017 22:46:41 +0000 (14:46 -0800)]
vkernel.7 - Update manual page

* Update the manual page with a better vkernel NFS client example.

7 years agovkernel: Make kptobj a static structure.
Antonio Huete Jimenez [Thu, 19 Jan 2017 08:03:16 +0000 (00:03 -0800)]
vkernel: Make kptobj a static structure.

  - There is a chicken-egg problem while allocating kptobj. Recent changes
    have made vm_object allocation to depend on kmalloc() which will call
    pmap_growkernel() when kptobj is not yet ready.

7 years agokernel: Add a new vm_object_init()
Antonio Huete Jimenez [Wed, 18 Jan 2017 23:59:27 +0000 (15:59 -0800)]
kernel: Add a new vm_object_init()

7 years agokernel: Rename vm_object_init() to vm_object_init1()
Antonio Huete Jimenez [Wed, 18 Jan 2017 22:37:31 +0000 (14:37 -0800)]
kernel: Rename vm_object_init() to vm_object_init1()

  - No functional change.

7 years agovkernel: Fix build
Antonio Huete Jimenez [Thu, 19 Jan 2017 08:02:58 +0000 (00:02 -0800)]
vkernel: Fix build

  - Partially-submitted by: zrj

7 years agolwp: vkernel actually uses some lwp_* functions
Antonio Huete Jimenez [Wed, 18 Jan 2017 22:19:42 +0000 (14:19 -0800)]
lwp: vkernel actually uses some lwp_* functions

7 years agosdhci - Simplify SDHCI_PRESENT_STATE register loop in sdhci_start_command().
Imre Vadász [Wed, 18 Jan 2017 20:34:52 +0000 (21:34 +0100)]
sdhci - Simplify SDHCI_PRESENT_STATE register loop in sdhci_start_command().

Taken-From: FreeBSD (svn r311693)

7 years agosdhci - Add a new interface method, get_card_present.
Imre Vadász [Wed, 18 Jan 2017 20:31:31 +0000 (21:31 +0100)]
sdhci - Add a new interface method, get_card_present.

Taken-From: FreeBSD (svn r311660)

7 years agoSync ACPICA with Intel's version 20170119.
Sascha Wildner [Thu, 19 Jan 2017 17:28:00 +0000 (18:28 +0100)]
Sync ACPICA with Intel's version 20170119.

Just some minor things:

* In iASL, fix an inadvertent remark.

* Remove the compiler bit width from the common signon of all tools.

* Add 2017 copyright.

For a more detailed list, please see sys/contrib/dev/acpica/changes.txt.

7 years agokernel/acpica: Include <stdarg.h> only for userland compilation.
Sascha Wildner [Thu, 19 Jan 2017 18:21:59 +0000 (19:21 +0100)]
kernel/acpica: Include <stdarg.h> only for userland compilation.

We generally don't include /usr/include/*.h headers in kernel code,
and the kernel doesn't need it.

It also caused issues when compiling with clang (due to redefinitions).

Reported-by: zrj
7 years agoUpdate the pciconf(8) database.
Sascha Wildner [Thu, 19 Jan 2017 17:49:01 +0000 (18:49 +0100)]
Update the pciconf(8) database.

January 8, 2017 snapshot from http://pciids.sourceforge.net/

7 years agocpdup: When verbose, also log the successful creation of a directory.
Sascha Wildner [Wed, 18 Jan 2017 17:44:30 +0000 (18:44 +0100)]
cpdup: When verbose, also log the successful creation of a directory.

7 years agoem/emx: Retry interrupt allocation with MSI
Sepherosa Ziehau [Wed, 18 Jan 2017 09:09:04 +0000 (17:09 +0800)]
em/emx: Retry interrupt allocation with MSI

Reported-by: jh32
7 years agotools/kq_sendrecv: Fix building on FreeBSD
Sepherosa Ziehau [Wed, 18 Jan 2017 05:46:30 +0000 (13:46 +0800)]
tools/kq_sendrecv: Fix building on FreeBSD

7 years agokernel - kmemstatistics list wasn't MPSAFE
Matthew Dillon [Tue, 17 Jan 2017 22:52:22 +0000 (14:52 -0800)]
kernel - kmemstatistics list wasn't MPSAFE

* Put a spinlock around the kmemstatistics list.  This list is modified
  whenever a kmalloc area is created or destroyed, and this generally does
  not happen concurrently except with kernels run by expert hackers like
  us and get scared.

* ...Or if a machine is running synth and/or doing a lot of concurrent
  tmpfs mounts and unmounts.

Reported-by: swildner
7 years agonpx.h: fix comment alignments (cosmetic)
John Marino [Tue, 17 Jan 2017 21:24:57 +0000 (15:24 -0600)]
npx.h: fix comment alignments (cosmetic)

7 years agoExpand fpusave union to include x86_64 versions
John Marino [Tue, 17 Jan 2017 20:00:55 +0000 (14:00 -0600)]
Expand fpusave union to include x86_64 versions

The structure of envmmx remains a constant size, but changes composition
between the i386 and x86_64 versions.  While here, flesh out the
832-byte saveymm structure, also distinguishing between 32 and 64 bit
versions.

While here, convert u_char, u_short, u_int16_t, u_int32_t to standard
integers; <sys/types.h> with it's redundant guards is no longer required
although <stdint> is now needed.

This allows the fpu context to be seen on x86_64 by casting the fpusave
union to the new savexmm64 and saveymm64 structures.

7 years agotools/kq_recvserv: Add option to set receiving thread's CPU affinity.
Sepherosa Ziehau [Tue, 17 Jan 2017 15:47:55 +0000 (23:47 +0800)]
tools/kq_recvserv: Add option to set receiving thread's CPU affinity.

7 years agotools/kq_recvserv: Add option to enable SO_REUSEPORT
Sepherosa Ziehau [Tue, 17 Jan 2017 15:27:41 +0000 (23:27 +0800)]
tools/kq_recvserv: Add option to enable SO_REUSEPORT

7 years agotools/kq_sendcli: Add option to set sending thread's CPU affinity
Sepherosa Ziehau [Tue, 17 Jan 2017 14:58:47 +0000 (22:58 +0800)]
tools/kq_sendcli: Add option to set sending thread's CPU affinity

7 years agotools/kq_recvserv: Add option to set read size
Sepherosa Ziehau [Tue, 17 Jan 2017 14:59:35 +0000 (22:59 +0800)]
tools/kq_recvserv: Add option to set read size

7 years agotool/kq_sendcli: Add option to set write size.
Sepherosa Ziehau [Tue, 17 Jan 2017 14:31:40 +0000 (22:31 +0800)]
tool/kq_sendcli: Add option to set write size.

7 years agotools/kq_sendcli: Add sendfile support.
Sepherosa Ziehau [Tue, 17 Jan 2017 14:18:46 +0000 (22:18 +0800)]
tools/kq_sendcli: Add sendfile support.

Based on work done by QianYue You.

7 years agosdhci - Use the right capability field for printing the bus width.
Imre Vadász [Mon, 16 Jan 2017 12:43:20 +0000 (13:43 +0100)]
sdhci - Use the right capability field for printing the bus width.

* In sdhci_init_slot(), use the right capability field for determining
  the announced bus width based on MMC_CAP_*_BIT_DATA.

Taken-From: FreeBSD (svn r311794)

7 years agolwp: Move all lwp syscalls to sys/lwp.h
Sepherosa Ziehau [Mon, 16 Jan 2017 14:04:28 +0000 (22:04 +0800)]
lwp: Move all lwp syscalls to sys/lwp.h

Discussed-with: swildner, zrj

7 years agoparam: Bump version according to the comment.
Sepherosa Ziehau [Mon, 16 Jan 2017 03:07:59 +0000 (11:07 +0800)]
param: Bump version according to the comment.

7 years agokernel/usb4bsd: Configure the correct bMaxPacketSize for control endpoints.
Imre Vadász [Sun, 15 Jan 2017 09:55:42 +0000 (10:55 +0100)]
kernel/usb4bsd: Configure the correct bMaxPacketSize for control endpoints.

* Configure the correct bMaxPacketSize for control endpoints before
  requesting the initial complete device descriptor and not as part of
  the subsequent babble error recovery. Babble means that the received
  USB packet was bigger than than configured maximum packet size. This
  only affects enumeration of FULL speed USB devices which use a
  bMaxPacketSize different from 8 bytes. This patch might help fix
  enumeration of USB devices which exhibit USB I/O errors in dmesg
  during boot.

Taken-From: FreeBSD (svn r295928)

7 years agoWipe the start of the disk when formatting the disk with MBR, too.
Sascha Wildner [Sun, 15 Jan 2017 19:41:22 +0000 (20:41 +0100)]
Wipe the start of the disk when formatting the disk with MBR, too.

7 years agoinstaller: Add basic UEFI system setup.
Sascha Wildner [Sun, 15 Jan 2017 19:32:06 +0000 (20:32 +0100)]
installer: Add basic UEFI system setup.

GPT partition 0 is a 128MB FAT32 EFI system partition. Microsoft
recommends at least 100MB and FAT32 in [1]. Partition 1 is a
"DragonFly Label64" (aka "dragonfly" in gpt(8)) for the rest of
the disk. It is analogous to slice 1 in our MBR setup, and has the
same name, e.g. /dev/da0s1. The rest of the UEFI setup is identical
to an MBR setup. disklabel64, newfs*, and so forth.

We currently only support installing DragonFly to the whole disk
in UEFI mode, and we don't set any NVRAM EFI variables during
installation, so for now we are copying our stage 1 loader to
\EFI\BOOT\BOOTX64.EFI on the EFI system partition.

Also, there isn't any support for Secure Boot.

[1] https://technet.microsoft.com/en-us/library/hh824839.aspx#DiskPartitionRules

7 years agoAdd a uefi(8) manual page giving some info on the UEFI boot procedure.
Sascha Wildner [Sun, 15 Jan 2017 14:01:29 +0000 (15:01 +0100)]
Add a uefi(8) manual page giving some info on the UEFI boot procedure.

Adapted-from: FreeBSD

7 years agorconfig: Add a script for setting up a HAMMER/UEFI system.
Sascha Wildner [Sun, 15 Jan 2017 14:00:48 +0000 (15:00 +0100)]
rconfig: Add a script for setting up a HAMMER/UEFI system.

7 years agonrelease: Add UEFI mode capable boot media.
Sascha Wildner [Sun, 15 Jan 2017 19:29:03 +0000 (20:29 +0100)]
nrelease: Add UEFI mode capable boot media.

Add an EFI system partition to the IMG (to boot on UEFI systems) but
keep MBR (to boot on legacy BIOS systems).

The approach in https://wiki.freebsd.org/UEFI#CD.2FDVD_Boot_under_UEFI
is used to make the ISO boot on both UEFI and legacy systems.

7 years agopthread: Implement pthread_attr_{get,set}affinity_np
Sepherosa Ziehau [Sun, 15 Jan 2017 08:53:39 +0000 (16:53 +0800)]
pthread: Implement pthread_attr_{get,set}affinity_np

7 years agox86_64: Fix machdep.smp_active sysctl type.
Sepherosa Ziehau [Sun, 15 Jan 2017 08:52:42 +0000 (16:52 +0800)]
x86_64: Fix machdep.smp_active sysctl type.

7 years agolwp: Add lwp_create2 syscll, which takes cpumask for scheduling.
Sepherosa Ziehau [Sat, 14 Jan 2017 13:16:08 +0000 (21:16 +0800)]
lwp: Add lwp_create2 syscll, which takes cpumask for scheduling.

7 years agoif_iwm - Configure the PCIe LTR, fix PCI express capability accesses.
Imre Vadász [Sun, 15 Jan 2017 09:42:49 +0000 (10:42 +0100)]
if_iwm - Configure the PCIe LTR, fix PCI express capability accesses.

Taken-From: Linux iwlwifi

7 years agokernel - Add PCIER_DEVCAP2 and PCIER_DEVCTRL2 definitions to pcireg.h.
Imre Vadász [Sun, 15 Jan 2017 09:43:31 +0000 (10:43 +0100)]
kernel - Add PCIER_DEVCAP2 and PCIER_DEVCTRL2 definitions to pcireg.h.

Partly Taken-From: FreeBSD

7 years agovgapci: Remove verbose printf() messages from pci_(get/set)_powerstate().
Imre Vadász [Sun, 15 Jan 2017 09:45:30 +0000 (10:45 +0100)]
vgapci: Remove verbose printf() messages from pci_(get/set)_powerstate().

7 years ago<stddef.h>: Add max_align_t guards for future reference.
zrj [Sat, 14 Jan 2017 09:39:35 +0000 (11:39 +0200)]
<stddef.h>: Add max_align_t guards for future reference.

There are some confusion how this c++11 feature should be implemented.

7 years agodrm/linux: Remove a redundant definition of NSEC_PER_SEC
François Tigeot [Sat, 14 Jan 2017 09:39:34 +0000 (10:39 +0100)]
drm/linux: Remove a redundant definition of NSEC_PER_SEC

7 years agodrm/linux: Add local_clock()
François Tigeot [Sat, 14 Jan 2017 09:34:35 +0000 (10:34 +0100)]
drm/linux: Add local_clock()

7 years agodrm/linux: Add acpi/video.h
François Tigeot [Sat, 14 Jan 2017 09:25:21 +0000 (10:25 +0100)]
drm/linux: Add acpi/video.h

7 years agoman: Add cross references
Sepherosa Ziehau [Sat, 14 Jan 2017 08:57:55 +0000 (16:57 +0800)]
man: Add cross references

7 years agopthread: Implement pthread_{set,get}affinity_np
Sepherosa Ziehau [Sat, 14 Jan 2017 08:12:40 +0000 (16:12 +0800)]
pthread: Implement pthread_{set,get}affinity_np

7 years agolibc: Take cpusetsize into consideration for sched_getaffinity()
Sepherosa Ziehau [Sat, 14 Jan 2017 08:11:50 +0000 (16:11 +0800)]
libc: Take cpusetsize into consideration for sched_getaffinity()

7 years agolibc: Add const to the cpu_set_t for sched_setaffinity.
Sepherosa Ziehau [Sat, 14 Jan 2017 06:39:36 +0000 (14:39 +0800)]
libc: Add const to the cpu_set_t for sched_setaffinity.

7 years agolibc: Add sched_{set,get}affinity().
Sepherosa Ziehau [Sat, 14 Jan 2017 02:18:55 +0000 (10:18 +0800)]
libc: Add sched_{set,get}affinity().

Same as other sched_ fucntions, these two functions do _not_ support
the overriding pid as tid as what Linux supports.

7 years agosecurity/libressl: Fix ECDSA P-256 timing attack vulnerability
John Marino [Fri, 13 Jan 2017 22:11:37 +0000 (16:11 -0600)]
security/libressl: Fix ECDSA P-256 timing attack vulnerability

OpenBSD 6.0 errata 16, Jan 5, 2017:
Avoid possible side-channel leak of ECDSA private keys when signing.

security: CVE-2016-7056

7 years agoUpdate libressl README.DRAGONFLY for v2.4.4
John Marino [Fri, 13 Jan 2017 22:11:28 +0000 (16:11 -0600)]
Update libressl README.DRAGONFLY for v2.4.4

7 years agoMerge branch 'vendor/LIBRESSL'
John Marino [Fri, 13 Jan 2017 23:15:53 +0000 (17:15 -0600)]
Merge branch 'vendor/LIBRESSL'

7 years agoUpdate LibreSSL from version 2.4.3 => 2.4.4
John Marino [Fri, 13 Jan 2017 22:00:47 +0000 (16:00 -0600)]
Update LibreSSL from version 2.4.3 => 2.4.4

7 years agoif_iwm - Update struct iwm_scan_results_notif. Remove old/unused definitions
Imre Vadász [Fri, 13 Jan 2017 17:00:45 +0000 (18:00 +0100)]
if_iwm - Update struct iwm_scan_results_notif. Remove old/unused definitions

* Remove outdated notifications IWM_SCAN_ABORT_CMD,
  IWM_SCAN_START_NOTIFICATION and IWM_SCAN_RESULTS_NOTIFICATION.

* Remove unused enum iwm_scan_complete_status.

* Use the updated FW Api version 3 of struct iwm_scan_results_notif.

* No functional change, since struct iwm_scan_results_notif is never
  accessed in iwm at the moment.

Taken-From: Linux iwlwifi commits 1083fd7391e989be52022f0f338e9dadc048b063
            and 75118fdb63496e4611ab50380499ddd62b9de69f.

7 years ago<sys/diskmbr.h>: Rename DOSPTYP_GPT to DOSPTYP_EFI.
Sascha Wildner [Thu, 12 Jan 2017 19:33:49 +0000 (20:33 +0100)]
<sys/diskmbr.h>: Rename DOSPTYP_GPT to DOSPTYP_EFI.

The former name was rather misleading since an EFI system partition can
also exist on an MBR disk (see 79363655e5db).

7 years agokernel: Fix GPT detection a bit.
Sascha Wildner [Thu, 12 Jan 2017 18:32:11 +0000 (19:32 +0100)]
kernel: Fix GPT detection a bit.

Having a legacy MBR partition of type 0xef aka EFI system
partition is a perfectly legal case, and it is covered by
the UEFI specification too (in version 2.6, see section 12.3.1).

That means it is not a sufficient indicator that this is a
disk with a GPT, so don't automatically treat it as one. For
the partition in the PMBR on actual GPT disks, type 0xee
will always be used, so that check alone is sufficient for
what is intended here.

Background: Using legacy MBR and having an EFI system
partition at the same time is the layout most USB installation
media use in order to boot on both UEFI and legacy BIOS systems.
The upcoming UEFI installation support work will do the same.

7 years agoacpi - Incidental MPSAFE (2)
Matthew Dillon [Wed, 11 Jan 2017 21:11:48 +0000 (13:11 -0800)]
acpi - Incidental MPSAFE (2)

* The changes in a639f788 revealed bubgs in acpi_thermal's support
  threads.  These threads start executing before the main acpi device
  finishes its attach.  The mplock serialization saved us before, and
  removing it revealed the problem.

* Force serialization using the main acpi lock to prevent the support
  threads from starting up until after the main acpi attach has
  completed.

Reported-by: swildner
7 years agokernel - Change vm.swapcache.use_chflags default to 0
Matthew Dillon [Wed, 11 Jan 2017 19:02:32 +0000 (11:02 -0800)]
kernel - Change vm.swapcache.use_chflags default to 0

* Change the vm.swapcache.use_chflags default to 0, meaning that swapcache
  will generally cache everything it sees if enabled.  Can be set back to
  1 by the sysop when more sophisticated operation is desired.

7 years agokernel - kernel - Incidental MPLOCK removal (usched, affinity)
Matthew Dillon [Wed, 11 Jan 2017 18:02:55 +0000 (10:02 -0800)]
kernel - kernel - Incidental MPLOCK removal (usched, affinity)

* Affinity code needs to be protected via p->p_token and lwp->lwp_token.
  Remove use of the mplock.

* If tid is -1 getaffinity() will lookup the lowest-numbered thread,
  and setaffinity will adjust ALL threads associated with the process.

* usched doesn't need mplock2.h

7 years agokernel - Incidental MPLOCK removal
Matthew Dillon [Wed, 11 Jan 2017 17:47:56 +0000 (09:47 -0800)]
kernel - Incidental MPLOCK removal

* Remove misc #include <sys/mplock2.h> statements that are no longer needed.

* Replace mplock with acct_lock in kern_acct.c

* Replace mplock with msg_token in sysv_msg.c

* Replace mplock with p->p_token in the profiling code.

7 years agousched: Fix the returned CPU affinity make by smp_active_mask.
Sepherosa Ziehau [Wed, 11 Jan 2017 16:09:51 +0000 (00:09 +0800)]
usched: Fix the returned CPU affinity make by smp_active_mask.

7 years agolwp: Add two syscalls to set/get lwp's CPU affinity mask.
Sepherosa Ziehau [Mon, 9 Jan 2017 15:48:18 +0000 (23:48 +0800)]
lwp: Add two syscalls to set/get lwp's CPU affinity mask.

7 years agousched: Implement LWP lazy migration support.
Sepherosa Ziehau [Wed, 11 Jan 2017 13:25:57 +0000 (21:25 +0800)]
usched: Implement LWP lazy migration support.

Non-self migration can be achieved by just setting LWP's cpumask,
which is almost costless.  The migration will happen upon next
time the target LWP is scheduled.

Reviewed-by: dillon@
7 years agokernel - Fix recursive module loading (from MPSAFE changes)
Matthew Dillon [Tue, 10 Jan 2017 21:49:00 +0000 (13:49 -0800)]
kernel - Fix recursive module loading (from MPSAFE changes)

* Fix recursive module loading which I broke in the
  recent MPSAFE work.

* Fix firmware loader callback deadlock.

* Fixes radeon drm lockup on startx

Reported-by: mneumann
7 years agodrm/i1915: Disable stolen memory support
François Tigeot [Tue, 10 Jan 2017 21:46:52 +0000 (22:46 +0100)]
drm/i1915: Disable stolen memory support

Some initialization code paths are not yet enabled, making
it crash-prone on some hardware

7 years agommcsd - Wasn't quite MPSAFE, fix startup race (2)
Matthew Dillon [Tue, 10 Jan 2017 03:11:47 +0000 (19:11 -0800)]
mmcsd - Wasn't quite MPSAFE, fix startup race (2)

* Needed to rearrange one more thing to report capacity properly.

7 years agommcsd - Wasn't quite MPSAFE, fix startup race
Matthew Dillon [Tue, 10 Jan 2017 03:08:03 +0000 (19:08 -0800)]
mmcsd - Wasn't quite MPSAFE, fix startup race

* mmcsd was attaching the drive before finishing the softc setup.  Being
  MPSAFE, the system tried to probe the mmcsd before it could finish the
  setup.

* Fixed by shifting things around a little.

7 years agokernel - Remove mplock in process trap/signal code (non-performance)
Matthew Dillon [Mon, 9 Jan 2017 23:23:33 +0000 (15:23 -0800)]
kernel - Remove mplock in process trap/signal code (non-performance)

* Remove the mplock from paths that no longer need it.  This is mostly
  incidental, since processes do not usually recover from traps like
  SIGBUS or SIGSEGV.

7 years agommcsd(4): Declare device D_MPSAFE. Move start_transaction to mmcsd thread.
Imre Vadász [Mon, 9 Jan 2017 22:00:42 +0000 (23:00 +0100)]
mmcsd(4): Declare device D_MPSAFE. Move start_transaction to mmcsd thread.

* SD card commands are issued in the mmcsd thread after it takes the bio
  command from the queue, so we should do the devstat_start_transaction()
  there, instead of directly in mmcsd_strategy().

7 years agokernel - Make misc crypto functions MPSAFE
Matthew Dillon [Mon, 9 Jan 2017 22:38:54 +0000 (14:38 -0800)]
kernel - Make misc crypto functions MPSAFE

* Remove mplock from the crypto device code.

* Implement cryptodev_lock for cryptof_ioctl() and related functions.

* Prevent cryptof functions from imploding on concurrent access.

7 years agobuf - Remove B_ORDERED, Incidental CAM MPSAFE (2)
Matthew Dillon [Mon, 9 Jan 2017 22:38:09 +0000 (14:38 -0800)]
buf - Remove B_ORDERED, Incidental CAM MPSAFE (2)

* Remove bowrite() operation in UFS (use bwrite() instead).

7 years agokernel - Incidental MPLOCK removal (devfs)
Matthew Dillon [Mon, 9 Jan 2017 22:36:07 +0000 (14:36 -0800)]
kernel - Incidental MPLOCK removal (devfs)

* Fix ordering issues for devfs initialization vs libprop that previously
  depended on the MPLOCK.

* Use a private devfs_token for other areas of devfs previously using
  the mplock.

* Make /dev/devfs MPSAFE.

7 years agokernel - Incidental MPLOCK removal
Matthew Dillon [Mon, 9 Jan 2017 22:34:11 +0000 (14:34 -0800)]
kernel - Incidental MPLOCK removal

* A few more mplocks in the moutn and umount paths.  Also fix a rootmount
  case (the latter being non-performance) for completeness.

7 years agokernel - Incidental MPLOCK removal (non-performance)
Matthew Dillon [Mon, 9 Jan 2017 22:28:42 +0000 (14:28 -0800)]
kernel - Incidental MPLOCK removal (non-performance)

* varsym is MPSAFE.

7 years agokernel - Incidental MPLOCK removal (non-performance)
Matthew Dillon [Mon, 9 Jan 2017 22:26:12 +0000 (14:26 -0800)]
kernel - Incidental MPLOCK removal (non-performance)

* proc filterops.

* kernel linkerops and kld code.

* Warn if a non-MPSAFE interrupt is installed.

* Use a private token in the disk messaging core (subr_disk) instead of
  the mp token.

* Use a private token for sysv shm adminstrative calls.

* Cleanup.

7 years agokernel - Remove global debug counters from device path
Matthew Dillon [Mon, 9 Jan 2017 22:22:59 +0000 (14:22 -0800)]
kernel - Remove global debug counters from device path

* Remove global debugging counters from the device path, including
  dev_dstrategy().  Improves performance when heavily concurrent I/O
  is being performed.

7 years agokernel - Incidental mplock removal - NTP support
Matthew Dillon [Mon, 9 Jan 2017 22:22:14 +0000 (14:22 -0800)]
kernel - Incidental mplock removal - NTP support

* Make the NTP support functions MPSAFE.

7 years agokernel - Incidental mplock removal
Matthew Dillon [Mon, 9 Jan 2017 22:20:13 +0000 (14:20 -0800)]
kernel - Incidental mplock removal

* We don't need to acquire the mplock when exec'ing the init program.

7 years agokernel - Make gpio and gpio_led MPSAFE
Matthew Dillon [Mon, 9 Jan 2017 22:03:46 +0000 (14:03 -0800)]
kernel - Make gpio and gpio_led MPSAFE

* Convert to a private lock.

7 years agoacpi - Incidental MPSAFE
Matthew Dillon [Mon, 9 Jan 2017 22:00:48 +0000 (14:00 -0800)]
acpi - Incidental MPSAFE

* Make ACPI devices MPSAFE (use a private token).

* Make the ACPI interupt MPSAFE.

7 years agobuf - Remove B_ORDERED, Incidental CAM MPSAFE
Matthew Dillon [Mon, 9 Jan 2017 21:58:04 +0000 (13:58 -0800)]
buf - Remove B_ORDERED, Incidental CAM MPSAFE

* Remove support for B_ORDERED.  Most drivers ignore it, and it is basically
  unusable.

* The CAM "xpt", "pass", and "sg" devices are MPSAFE, flag them
  appropriately.

7 years agokernel - Incidental MPLOCK removal (non-performance)
Matthew Dillon [Mon, 9 Jan 2017 21:54:05 +0000 (13:54 -0800)]
kernel - Incidental MPLOCK removal (non-performance)

* Remove the MPLOCK in a ton of places that don't need it or can be
  trivially tokenized or locked.

* Mostly non-performance adjustments, though the X server does use
  scmouse or sysmouse.

7 years agosdhci - Use MPSAFE version for the taskqueue and callouts
Matthew Dillon [Mon, 9 Jan 2017 21:47:38 +0000 (13:47 -0800)]
sdhci - Use MPSAFE version for the taskqueue and callouts

* Use the MPSAFE version for the taskqueue and callouts.

7 years agodrm/i915: Revert DisplayPort fast link training feature
François Tigeot [Mon, 9 Jan 2017 22:31:14 +0000 (23:31 +0100)]
drm/i915: Revert DisplayPort fast link training feature

DisplayPort fast link training causes screen flickering on some hw combinations

Obtained-from: Linux 4.6.7, Commit b0f7095f8bd503eddf9a91a258e796da0058f182

7 years agodrm/i915/psr: Try to program link training times correctly
François Tigeot [Mon, 9 Jan 2017 22:22:54 +0000 (23:22 +0100)]
drm/i915/psr: Try to program link training times correctly

Increase link training time to 2.5ms, fixing
flickering issues associated with PSR.

Obtained-from: Linux 4.6.7, commit 69205c5c16df0702870179c8419e89cc954bd311

7 years agosdhci - Fix typo, support bits for drive types A,C,D are in capabilities.
Imre Vadász [Fri, 6 Jan 2017 21:47:28 +0000 (22:47 +0100)]
sdhci - Fix typo, support bits for drive types A,C,D are in capabilities.

* The drive type support flags in the capabilities register (0x40) are for
  drive types A,C,D, drive type B is the default setting (value 0) of the
  drive strength field in the HOST_CONTROL2 register.

7 years agosdhci - Add some missing register constants, and dump some more registers.
Imre Vadász [Thu, 5 Jan 2017 20:17:49 +0000 (21:17 +0100)]
sdhci - Add some missing register constants, and dump some more registers.

* Adds defines for content of SDHCI_ADMA_ERR and SDHCI_HOST_CONTROL2
  registers.

* Add slot type capability bits. These bits should allow recognizing
  removable card slots, embedded cards and shared buses (shared bus
  supposedly is always comprised of non-removable cards).

* Dump CAPABILITIES2, ADMA_ERR, HOST_CONTROL2 and ADMA_ADDRESS_LOW
  registers in sdhci_dumpregs().

7 years agosdhci - Use bus_dmamem_coherent for allocating memory for sdma DMA.
Imre Vadász [Thu, 5 Jan 2017 18:34:24 +0000 (19:34 +0100)]
sdhci - Use bus_dmamem_coherent for allocating memory for sdma DMA.

* While there, rename various constants which are related to SDMA (and not
  going to be used for ADMA2) from *_DMA_* names to *_SDMA_*.

7 years agousched: Add USCHED_SET_CPUMASK.
Sepherosa Ziehau [Mon, 9 Jan 2017 13:33:30 +0000 (21:33 +0800)]
usched: Add USCHED_SET_CPUMASK.

7 years agotest/passdesc: Allow specifying the payload length.
Sepherosa Ziehau [Mon, 9 Jan 2017 05:03:57 +0000 (13:03 +0800)]
test/passdesc: Allow specifying the payload length.

We don't suffer the following bug though.

Inspired-by: FreeBSD https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=181741
7 years agokernel - Further refactor vmstats, adjust page coloring algorithm
Matthew Dillon [Mon, 9 Jan 2017 04:03:38 +0000 (20:03 -0800)]
kernel - Further refactor vmstats, adjust page coloring algorithm

* Further refactor vmstats by tracking adjustments in gd->gd_vmstats_adj
  and doing a copyback of the global vmstats into gd->gd_vmstats.  All
  code critical paths access the localized copy to test VM state, removing
  most global cache ping pongs of the global structure.   The global
  structure 'vmstats' still contains the master copy.

* Bump PQ_L2_SIZE up to 512.  We use this to localized the VM page queues.
  Make some adjustments to the pg_color calculation to reduce (in fact
  almost eliminate) SMP conflicts on the vm_page_queue[] between cpus
  when the VM system is operating normally (not paging).

* This pumps the 4-socket opteron test system up to ~4.5-4.7M page
  faults/sec in testing (using a mmap/bzero/munmap loop on 16MB x N
  processes).

  This pumps the 2-socket xeon test system up to 4.6M page faults/sec
  with 32 threads (250K/sec on one core, 1M on 4 cores, 4M on 16 cores,
  5.6M on 32 threads).  This is near the theoretical maximum possible for
  this test.

* In this particular page fault test, PC sampling indicates *NO* further
  globals are undergoing cache ping-ponging.  The PC sampling predominantly
  indicates pagezero(), which is expected.  The Xeon is zeroing an aggregate
  of 22GBytes/sec at 32 threads running normal vm_fault's.

7 years agokernel - Refactor struct vmstats and vm_zone
Matthew Dillon [Mon, 9 Jan 2017 00:02:59 +0000 (16:02 -0800)]
kernel - Refactor struct vmstats and vm_zone

* These changes significantly improve the simultaneous non-conflicting
  VM fault rate.  On our 4-socket opteron (48 cores, which makes a great
  test case because its cache mastership stalls are so expensive), the
  maximum concurrent VM fault rate increased from ~2.4M/sec to ~3.5M/sec,
  and suffers no degredation after topping out.

* Refactor the fields in struct vmstats to separate out mostly
  read-only variables from nominally modified variables, reducing
  cache mastership stalls.

* Remove vm_shared_hit, vm_shared_count, and vm_shared_miss sysctl
  statistics, removing related cache mastership stalls from the
  critical path.

* Move the spinlock in vpgqueues to the base of the structure.

* Increase the vmstats slop (how large a negative value can accumulate
  in pcpu stats before rolling it up).

* Fix cache mastership stalls in the zalloc() and zfree() paths by
  consolidating pcpus elements into its own cache-aligned structure
  and giving each pcpu its on znalloc counter.