Displaying 1 50 of 35,897 commits (0.009s)

DragonFlyBSD — sys/netinet tcp_output.c

tcp: Don't prematurely drop receiving-only connections.

If the connection was persistent and receiving-only, several (12)
sporadic device insufficient buffers would cause the connection be
dropped prematurely:
Upon ENOBUFS in tcp_output() for an ACK, retransmission timer is
started.  No one will stop this retransmission timer for receiving-
only connection, so the retransmission timer promises to expire and
t_rxtshift is promised to be increased.  And t_rxtshift will not be
reset to 0, since no RTT measurement will be done for receiving-only
connection.  If this receiving-only connection lived long enough,
and it suffered 12 sporadic device insufficient buffers, i.e.
t_rxtshift >= 12, this receiving-only connection would be dropped
prematurely by the retransmission timer.

We now assert that for data segments, SYNs or FINs either rexmit or
persist timer was wired upon ENOBUFS.  And don't set rexmit timer
for other cases, i.e. ENOBUFS upon ACKs.

And we no longer penalize send window upon ENOBUFS.

Obtained-from: FreeBSD r300981
Delta File
+4 -18 sys/netinet/tcp_output.c
+4 -18 1 file

DragonFlyBSD — share/man/man9 ieee80211.9

ieee80211.9: Fix .Nm
Delta File
+1 -1 share/man/man9/ieee80211.9
+1 -1 1 file

DragonFlyBSD — secure/lib/libcrypto/man CMS_sign_add1_signer.3

Remove an unused manual page.

DragonFlyBSD — sys/platform/pc64/x86_64 pmap.c

kernel - Align virtual_start, workaround possible HW VM bugs

* virtual_start is 2MB-aligned at the kernel load's end address.
  The page table mappings prior to this address are 2MB mappings.

  Additional SYSMAP()d space is reserved in early boot, such as
  the CMAP entries and the dmesg buffer.  The page table mappings
  in this area should all be 4K pages.

* Kernel virtual memory beyond the SYSMAP()d space is available for
  general allocation, but positioned such that only kldload modules will
  try to use it specifically.  Sascha (swildner) found that any initial
  kldload done after boot in vmware player 7.1.4 build-3848939 running
  on a i7-4770K based host would panic.

  A kernel core of this panic appears to show a corrupt pte in a location
  where the pte should have been read as 0.

* Doing an additional 2MB alignment of virtual_start after the SYSMAP()s
  appears to work around this corruption, which is this commit.  This
  means that the first general allocation in the kernel load portion of
  the kernel_map will be shifted to the next 2MB boundary past the SYSMAP()s.

* I can only speculate that the hardware is somehow getting confused, but
  I can't really fathon how it is getting confused because the SYSMAP
  reservations should all be using 4KB pages.
Delta File
+1 -0 sys/platform/pc64/x86_64/pmap.c
+1 -0 1 file

DragonFlyBSD — sys/bus/gpio/gpio_acpi gpio_acpi.c

gpio_acpi: Factor out gpio_acpi_check_gpioint().

* gpio_acpi_check_gpioint() can be reused for the device-driver facing
  interface for dealing with gpio acpi-resources.
Delta File
+42 -30 sys/bus/gpio/gpio_acpi/gpio_acpi.c
+42 -30 1 file

DragonFlyBSD — sys/bus/gpio/gpio_acpi gpio_acpi.c

gpio_acpi: Factor out gpio_acpi_check_gpioint().

* gpio_acpi_check_gpioint() can be reused for the device-driver facing
  interface for dealing with gpio acpi-resources.
Delta File
+42 -30 sys/bus/gpio/gpio_acpi/gpio_acpi.c
+42 -30 1 file

DragonFlyBSD — sys/boot/dloader loader-bootp.conf loader.conf, sys/config X86_64_GENERIC LINT64

Some more cleanup in comments etc. after the an(4) removal.

DragonFlyBSD — share/man/man9 ieee80211_vap.9

ieee80211_vap.9: Fix date.
Delta File
+1 -1 share/man/man9/ieee80211_vap.9
+1 -1 1 file

DragonFlyBSD — . Makefile.inc1, share/man/man4 drm.4

Various typo and wording fixes.

DragonFlyBSD — sys/dev/virtual/virtio/virtio virtqueue.c

virtio - fix contigmalloc muff last commit

* Have to use M_WAITOK for contigmalloc, not M_INTWAIT.

Reported-by: tuxillo

DragonFlyBSD — sys/dev/virtual/virtio/block virtio_blk.c

virtio - vtblk_finish_bio() called with serializer held

* vtblk_finish_bio() was being improperly called with the
  serializer held in the strategy error path.

* Note that biodone() is properly called without the
  serializer held in the normal completion path.

DragonFlyBSD — sys/contrib/dev/acpica changes.txt, sys/contrib/dev/acpica/source/compiler asloptions.c

Sync ACPICA with Intel's version 20160527.

* Temporarily reverted the new arbitrary bit length/alignment support
  in AcpiHwRead/AcpiHwWrite for the Generic Address Structure. There
  have been a number of regressions with the new code that need to be
  fully resolved and tested before this support can be finally
  integrated into ACPICA. Apologies for any inconveniences these
  issues may have caused.

* The ACPI message macros are not configurable (ACPI_MSG_ERROR,
  ACPI_MSG_EXCEPTION, ACPI_MSG_WARNING, ACPI_MSG_INFO,
  ACPI_MSG_BIOS_ERROR, and ACPI_MSG_BIOS_WARNING). Lv Zheng.

* Fixed a couple of GCC warnings associated with the use of the
  -Wcast-qual option. Adds a new return macro, return_STR.
  Jung-uk Kim.

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

DragonFlyBSD — sys/dev/virtual/virtio/block virtio_blk.c, sys/dev/virtual/virtio/virtio virtqueue.c virtio.c

virtio - Fix a few porting issues and bugs

* vtblk_maximum_segments() was really buggy.  It was
  adding the base min segments to blkcfg->seg_max,
  miscalculating values for the case where the F_SEG_MAX
  feature is not in the feature list.

  This could result in virtio imploding on a too-small
  sglist.

* Fix situations where M_NOWAIT is improperly used.

* At least use 16-byte alignment for contigmalloc()'s.

DragonFlyBSD — share/man/man4 wlan.4

wlan.4: Comment out reference to bwi(4) which we do not have.
Delta File
+1 -1 share/man/man4/wlan.4
+1 -1 1 file

DragonFlyBSD — share/man/man4 wlan.4 wlan_amrr.4, sys/contrib/dev/iwi Makefile

Various tiny updates in wlan*.4 manual pages and other related areas.

DragonFlyBSD — share/man/man9 ieee80211_beacon.9 ieee80211.9

ieee80211*.9: Fix some prototypes in the manual pages.

DragonFlyBSD — sys/platform/pc64/x86_64 pmap.c

kernel - Augment pte assertion with additional info

* Augment the invalid PTE assertion to print additional info.  This bug was
  thought to have been squashed but we received a recent report of it still
  occurring (on an older cpu).
Delta File
+6 -4 sys/platform/pc64/x86_64/pmap.c
+6 -4 1 file

DragonFlyBSD — share/man/man4 wi.4 iwnfw.4

Sync wlan driver manual pages with FreeBSD.
Delta File
+176 -167 share/man/man4/wi.4
+19 -7 share/man/man4/iwnfw.4
+11 -11 share/man/man4/ath.4
+18 -4 share/man/man4/iwn.4
+16 -4 share/man/man4/bwn.4
+15 -4 share/man/man4/ral.4
+28 -22 9 files not shown
+283 -219 15 files

DragonFlyBSD — sys/kern uipc_msg.c, sys/net netmsg.h

tcp: Make setsockopt(2) TCP_{NODELAY,NOPUSH,NOOPT,FASTKEEP} asynchronous

This makes all network system calls used by nginx asynchronous and greatly
improves nginx performance.

For 1 requests/connection (30K concurrent connections), this commit gives
12% improvement for 16KB web object (was 16.8Gbps, now 19.2Gbps, which
maxes out 2x10Ge).  And it gives 19% improvement for 1KB web object (was
156Kreqs/s, now 186Kreqs/s).

DragonFlyBSD — etc/devd usb.conf

Regenerate devd's usb.conf for accumulated usb driver updates.
Delta File
+75 -7 etc/devd/usb.conf
+75 -7 1 file

DragonFlyBSD — sys/dev/netif/iwm if_iwm.c

if_iwm - valid_{tx,rx}_ant from radio_cfg is only needed for 8000 family.

* The "if (!data->valid_tx_ant || !data->valid_rx_ant) {" check was getting
  triggered with a 3165 chipset.
Delta File
+0 -10 sys/dev/netif/iwm/if_iwm.c
+0 -10 1 file

DragonFlyBSD — . Makefile_upgrade.inc, share/man/man4 ieee80211.4 Makefile

Sync our ieee80211* manpages with FreeBSD (plus some adjustments by me).

DragonFlyBSD — sys/dev/netif/ral rt2560.c rt2661.c

ral - Add missing BUS_DMA_NOWAIT to bus_dmamap_load_mbuf_segment() calls

* Add missing BUS_DMA_NOWAIT to bus_dmamap_load_mbuf_segment() calls.

* NOTE: FreeBSD original code in #else not modified, but probably needs
  the same change (to bus_dmamap_load_mbuf_sg()).

DragonFlyBSD — share/man/man4 ieee80211.4 wi.4, usr.sbin/802_11/wpa_supplicant wpa_supplicant.8

Remove an(4)/ancontrol(8) references from several manual pages.

DragonFlyBSD — sys/config LINT64

Add missing LINT64 options ({IWN,WPI}_DEBUG, URTWN_WITHOUT_UCODE).
Delta File
+3 -0 sys/config/LINT64
+3 -0 1 file

DragonFlyBSD — sys/dev/netif/wpi if_wpi_debug.h if_wpi.c

kernel/wpi: Fix build with WPI_DEBUG.

DragonFlyBSD — sys/dev/netif/ath/ath_hal ah_osdep.c

kernel/ath_hal: Add missing __printflike().

DragonFlyBSD — sys/conf options, sys/config LINT64

kernel: Remove two unused options, AH_REGOPS_FUNC, IEEE80211_AMPDU_AGE.

DragonFlyBSD — sys/vfs/hammer hammer_blockmap.c

sys/vfs/hammer: Remove unnecessary =NULL;

resx is only used (or should be declared) in else{} scope.
Delta File
+1 -3 sys/vfs/hammer/hammer_blockmap.c
+1 -3 1 file

DragonFlyBSD — sys/vfs/hammer hammer_vfsops.c

sys/vfs/hammer: Don't store negative free/avail for statfs/statvfs

f_bfree and f_bavail (these two equal on HAMMER) are likely to
be negative when the filesystem usage has reached its limit or
almost reaches limit.

There is no point in having negative values by using internally
reserved size (means not physically finalized with actual data),
so force 0 instead of -.

Note that this doesn't fix ENOSPC issues.

-- before this commit
 # df -T /HAMMER
 Filesystem  Type   1K-blocks      Used Avail Capacity  Mounted on
 TEST        hammer 116359168 116363168 -4000   100%    /HAMMER

-- with this commit
 # df -T /HAMMER
 Filesystem  Type   1K-blocks      Used Avail Capacity  Mounted on
 TEST        hammer 116359168 116359168     0   100%    /HAMMER
Delta File
+4 -0 sys/vfs/hammer/hammer_vfsops.c
+4 -0 1 file

DragonFlyBSD — sys/vfs/hammer hammer_blockmap.c hammer_object.c

sys/vfs/hammer: Add warning messages on ENOSPC

Add warning messages if the blockmap allocator finds ENOSPC.

B-Tree functions may also return ENOSPC, but those are results
of failure in blockmap allocation on node split.

_hammer_checkspace() doesn't have this warning for now.
This gets called by many of the syscalls before inmemory/ondisk
file operations start, so it's pretty obvious without bunch of
"No space left" in dmesg.

DragonFlyBSD — sys/vfs/hammer hammer_vfsops.c hammer.h

sys/vfs/hammer: Remove vfs.hammer.count_extra_space_used

This was added in a7e9bef1 in 2008 for debugging.
Exposing this to userspace isn't useful for users anyway considering
this is a sum of several internal parameters on resource usage.

DragonFlyBSD — . Makefile_upgrade.inc

make upgrade: Remove the if_an.ko module too (while here, sort a bit).
Delta File
+2 -1 Makefile_upgrade.inc
+2 -1 1 file

DragonFlyBSD — . UPDATING

Mention that an(4) has been removed in UPDATING.
Delta File
+13 -6 UPDATING
+13 -6 1 file

DragonFlyBSD — sys/netproto/802_11/wlan ieee80211_dragonfly.c

kernel/ieee80211: Remove an unused malloc type.

DragonFlyBSD — sys/dev/netif/iwn if_iwn_debug.h

kernel/iwn: Fix building with IWN_DEBUG.
Delta File
+1 -1 sys/dev/netif/iwn/if_iwn_debug.h
+1 -1 1 file

DragonFlyBSD — sys/dev/netif/bwn/bwn if_bwn_debug.h, sys/dev/netif/bwn/siba sibavar.h

kernel/bwn: Remove an unnecessary #define and add siba_debug.

DragonFlyBSD — sys/dev/netif/bwn/siba siba_core.c

kernel/bwn/siba: Fix a struct member name.

DragonFlyBSD — sys/bus/u4b/wlan if_urtwn.c

kernel/urtwn: Fix building with USB_DEBUG.
Delta File
+7 -7 sys/bus/u4b/wlan/if_urtwn.c
+7 -7 1 file

DragonFlyBSD — sys/dev/netif/ath/ath if_ath_tx.c

kernel/ath: Fix a misplaced #endif
Delta File
+1 -2 sys/dev/netif/ath/ath/if_ath_tx.c
+1 -2 1 file

DragonFlyBSD — sys/dev/netif/ath/ath if_ath_ioctl.c

kernel/ath: Fix building with ATH_DIAGAPI.

DragonFlyBSD — sys/contrib/dev/ath/ath_hal/ar9300 ar9300_xmit.c ar9300_gpio.c

kernel/ath: Fix building with AH_DEBUG.

DragonFlyBSD — sys/conf files

wlan - Add missing kernel config files for ath.

* Add missing files for 'ath' to the kernel build (for building
  ath into the kernel).
Delta File
+5 -0 sys/conf/files
+5 -0 1 file

DragonFlyBSD — sys/dev/netif/wpi if_wpivar.h

wlan - Fix lockstatus debugging

* Fix lockstatus debugging (was missing 'curthread' arg).
Delta File
+3 -3 sys/dev/netif/wpi/if_wpivar.h
+3 -3 1 file

DragonFlyBSD — sys/conf options, sys/dev/netif/wpi Makefile

wlan - Add missing options conf

* Add missing options conf for opt_wpi.h.  Also add opt_iwn.h in case we
  need it in the future (FreeBSD also had it).  We already had opt_iwm.h.
Delta File
+7 -0 sys/dev/netif/wpi/Makefile
+6 -0 sys/conf/options
+13 -0 2 files

DragonFlyBSD — sys/dev/netif/iwn if_iwn.c

iwn - Use pci_alloc_1intr()

* Use pci_alloc_1intr() in DragonFly for conditionalizing MSI interrupt
  installation.
Delta File
+7 -2 sys/dev/netif/iwn/if_iwn.c
+7 -2 1 file

DragonFlyBSD — sys/config X86_64_GENERIC, sys/dev/netif Makefile

wifi - Restore kernel build

* Restore wpi in netif Makefile and restore X86_64_CONFIG for all
  drivers originally in the kerne build, except 'an'.
Delta File
+10 -10 sys/config/X86_64_GENERIC
+1 -2 sys/dev/netif/Makefile
+11 -12 2 files

DragonFlyBSD — sys/contrib/dev/wpi LICENSE, sys/dev/netif/wpi if_wpi.c if_wpireg.h

wlan - Sync dev/netif/wpi from FreeBSD

* Sync dev/netif/wpi from FreeBSD, fbsd git dd885b9a0a0e, May 11 2016.

* Straight re-port from FreeBSD using new infrastructure, with DragonFly
  mods.

DragonFlyBSD — sys/dev/drm drm_dragonfly.c drm_drv.c, sys/dev/drm/i915 i915_dma.c

drm: Fix missing revision init.

While there, make i915 irq_install more consistent with drm.

Lets see what breaks.