DragonFlyBSD/src 0953052sys/conf files, sys/dev/drm/drm Makefile

drm/ttm: Remove two useless files

DragonFlyBSD/src a7f15a6sys/dev/drm/include/asm-generic atomic-long.h, sys/dev/drm/include/linux atomic.h

drm/linux: Add atomic_long_inc_not_zero()

Obtained-from: FreeBSD

DragonFlyBSD/src 2bbc773contrib/gcc-4.7/gcc gcc.c

gcc47: Add local -save-temps=objects option.

Just for convenience, same as in:
df642abc046981dfea4020a80f466a1acc7607ca

DragonFlyBSD/src df642abcontrib/gcc-8.0 README.DRAGONFLY, contrib/gcc-8.0/gcc gcc.c

gcc80: Add local -save-temps=objects option.

The gcc currently supports -save-temps= cwd, obj, object(not documented).
None of those variants can be used at all during buildworld sequence,
even at serial buildworld, without interfering with build in funny ways.
This is very easy to notice while compiling libraries, where rules are:
 * foo.c --> foo.o  (non-pic)
 * foo.c --> foo.So (pic)
 * foo.c --> foo.po (profiled)

The added option forces to use full object name (if it was supplied)
when constructing names for .s and .i intermediates, as follows:
 cc -c foo.c -o foo.o --> foo.o.s + foo.o.i
This allows to use -save-temps globally during buildworld and it is
*less* likely to interfere while producing normally created objects.
Repeated quickworld likely *will* have side effects, because the '-o'
flags are not used for kernel and programs (foo.s will shadow foo.c).

Tested by binary comparing /usr/obj trees for differences.

DragonFlyBSD/src 59950cblib/csu Makefile

csu: Expand .note section manipulation.

No functional change.
DeltaFile
+3-0lib/csu/Makefile
+3-01 files

DragonFlyBSD/src fba5272contrib/smbfs README, contrib/smbfs/lib/smb subr.c kiconv.c

smbfs - Bring in some string safety and type-o corrections

* Bring in some string safety fixes and type-o corrections from
  FreeBSD.

Submitted-by: Varun C H
Taken-from: FreeBSD

DragonFlyBSD/src 402014dcontrib/smbfs README, contrib/smbfs/lib/smb subr.c nbns_rq.c

smbfs - Bring in some string safety and type-o corrections

* Bring in some string safety fixes and type-o corrections from
  FreeBSD.

Submitted-by: Varun C H
Taken-from: FreeBSD

DragonFlyBSD/src 6adc224lib/libthread_xu/thread thr_syscalls.c

pthreads - sleep() needs strong reference

* Fix problem with sleep() not canceling due to missing strong
  reference.

Reported-by: godfrey

DragonFlyBSD/src 811de3blib/libthread_xu/thread thr_syscalls.c

pthreads - sleep() needs strong reference

* Fix problem with sleep() not canceling due to missing strong
  reference.

Reported-by: godfrey

DragonFlyBSD/src 4864151usr.bin/ministat ministat.c

ministat(1): Fix terminal width handling

ministat(1) will prefer to get the terminal width from the "COLUMNS"
environment variable, in which case the plot will overflow by 2 columns
because the plot frame is not considered.  Fix this issue.

However, in Emacs's "shell" mode (not the "term" mode), the plot will
still overflow because Emacs puts a '\' at the right-most column.  One
can manually modify the "COLUMNS" environment variable or use the "-w"
option to workaround this issue.

Reported-by: swildner

DragonFlyBSD/src ddb4155sys/vfs/hammer2 hammer2_vfsops.c

hammer2 - Fix panic on corrupt root inode on mount

* If the root inode is corrupt at mount time H2 was panicing
  instead of cleanly aborting.  Fixed by adding missing initialization.

* This case can occur when the device is partially overwritten by
  a dd or another filesystem.  Redundant volume headers may still alow
  H2 to find one, but the iroot might reside at a lower block number and
  already have been blown away.

Reported-by: BJoe / Bug #3181

DragonFlyBSD/src 5c88385sys/vfs/hammer2 hammer2_vfsops.c

hammer2 - Fix panic on corrupt root inode on mount

* If the root inode is corrupt at mount time H2 was panicing
  instead of cleanly aborting.  Fixed by adding missing initialization.

* This case can occur when the device is partially overwritten by
  a dd or another filesystem.  Redundant volume headers may still alow
  H2 to find one, but the iroot might reside at a lower block number and
  already have been blown away.

Reported-by: BJoe / Bug #3181

DragonFlyBSD/src d0fba32. Makefile_upgrade.inc

Remove a deleted header via 'make upgrade'.

DragonFlyBSD/src 7c7505dcontrib/gcc-5.0/gcc dwarf2out.c, contrib/gcc-5.0/gcc/config/i386 i386.c i386.md

gcc50: Retire sources and makefiles.

DragonFlyBSD/src 92a4261. Makefile.inc1 Makefile_upgrade.inc, etc/mtree BSD.include.dist BSD.usr.dist

gcc50: Disconnect from buildworld.

You served us good.

DragonFlyBSD/src 4b2fb98lib/libkvm kvm.3

kvm.3: More markup. Also begin sentences on a new line.
DeltaFile
+15-5lib/libkvm/kvm.3
+15-51 files

DragonFlyBSD/src 19cf04alib/libkvm kvm.3

kvm.3: Remove unnecessary argument to .Nm
DeltaFile
+1-1lib/libkvm/kvm.3
+1-11 files

DragonFlyBSD/src 421e68dlib/libc/sys wait.2

libc - Update manual page for wait.2

* Update the manual page for wait.2, properly describing the
  si_code and si_status fields of the siginfo_t structure to
  match recent fixes.
DeltaFile
+24-5lib/libc/sys/wait.2
+24-51 files

DragonFlyBSD/src c3a64adsys/kern kern_exit.c

kernel - Fix info->si_code for waitid() (2)

* Our si_status was W*() encoded, but linux does not W*() encode
  theirs.

  Do not W*() encode ours either.  Also properly decode the exit signal
  and exit code.

DragonFlyBSD/src 819542dsys/kern kern_exit.c

kernel - Fix info->si_code for waitid() (2)

* Our si_status was W*() encoded, but linux does not W*() encode
  theirs.

  Do not W*() encode ours either.  Also properly decode the exit signal
  and exit code.

DragonFlyBSD/src 3812af4sys/kern kern_exit.c

kernel - Fix info->si_code for waitid()

* info->si_code was improperly setting CLD_KILLED for normal exit()s
  with non-zero exit codes.

Reported-by: tuxillo

DragonFlyBSD/src bfd192fsys/kern kern_exit.c

kernel - Fix info->si_code for waitid()

* info->si_code was improperly setting CLD_KILLED for normal exit()s
  with non-zero exit codes.

Reported-by: tuxillo

DragonFlyBSD/src 82391b5etc man.conf, usr.bin/man man.c man.conf.5

Remove the unhooked/-used man(1) command from usr.bin.

It was added in 2013 but not continued on. I will soon replace the man(1)
we use (in gnu/usr.bin) with FreeBSD's.

DragonFlyBSD/src 99fecc7sys/dev/video/bktr bktr_i2c.h

kernel: Remove a no longer needed header file.

DragonFlyBSD/src afcaf2e. Makefile_upgrade.inc, share/misc man.template Makefile

Remove /usr/share/misc/man.template.

We don't write new manpages with these macros.

DragonFlyBSD/src 3185d15sys/dev/drm/ttm ttm_bo_manager.c

drm/ttm: Sync ttm_bo_man_put_node() with Linux 3.18

DragonFlyBSD/src c5fe25fgnu/usr.bin/rcs/rcs rcsfile.5

rcsfile.5: Fix mandoc -Wunsupp warning.

DragonFlyBSD/src 989f740contrib/mdocml lib.in, usr.bin/mandoc Makefile

mandoc(1): Further adjustments.

* Add some libraries we reference with .Lb.

* Add mandoc(1)'s makewhatis.8 to the yet unused manual pages.

DragonFlyBSD/src d6e3d4fcontrib/mdocml README.DELETED README.DRAGONFLY, usr.bin/mandoc config.h Makefile

Local adjustments for the mandoc(1) upgrade.

DragonFlyBSD/src a5dca70contrib/mdocml roff.c mdoc_validate.c

Merge branch 'vendor/MDOCML'

DragonFlyBSD/src 54ba960contrib/mdocml roff.c mdoc_validate.c

Import mandoc-1.14.5 and leave only the files we need.

It is in preparation for eventually switching to it for manual page
displaying.

DragonFlyBSD/src 1dac82asbin/gpt show.c, share/man/man7 tuning.7

Fix some typos in manpages/messages.

DragonFlyBSD/src 590d925share/misc pci_vendors

Update the pciconf(8) database.

March 14, 2019 snapshot from https://pci-ids.ucw.cz
DeltaFile
+108-24share/misc/pci_vendors
+108-241 files

DragonFlyBSD/src e404160share/man/man9 idr.9

idr.9: Fix idr_remove() prototype.

DragonFlyBSD/src 252524esys/dev/drm/include/linux wait.h

drm/linux: Improve linux/wait.h

DragonFlyBSD/src 63aa39bsys/dev/drm/include/asm atomic.h

drm/linux: Add ATOMIC_INIT()

DragonFlyBSD/src 3851e4busr.bin/ministat ministat.c ministat.1

Import ministat(1) from FreeBSD

This utility calculates fundamental statistical properties of the input
data and determines whether or not there is statistical difference
between them.

From the HISTORY section of the man page:
"The ministat command was written by Poul-Henning Kamp out of
frustration over all the bogus benchmark claims made by people with no
understanding of the importance of uncertainty and statistics."

DragonFlyBSD/src 70839a9. Makefile_upgrade.inc, libexec/rtld-elf rtld.1 Makefile

i386 removal, part 71/x: Remove legacy FreeBSD brand.

We had a compat layer in i386 that was intended to run old DragonFly
binaries from <1.2 times if my memory serves well. That also worked
with FreeBSD 4.x binaries. It was removed in part 7 and 8 of this
commit series.

This commit reverts cee191cabfed209fdad9972bd88be5b18e310361.

It also removes all references to ld-elf.so.1.

Reported-by: zrj

DragonFlyBSD/src a34b416sys/dev/drm drm_mm.c, sys/dev/drm/include/drm/ttm ttm_bo_api.h ttm_bo_driver.h

drm/ttm: convert to unified vma offset manager

* Add TTM_BO_PRIV_FLAG_ACTIVE.

* Changes so we can use more of the linux ttm_bo_vm.c code.  Fake
  struct vm_fault, fake struct vm_operations_struct, and adjust
  struct vm_area_struct.  Also add related flags.

  The VM interface to DragonFlyBSD is now (mostly) a wrapper around
  the linux code.

* The linux red/black tree code uses some of the DFly RB macros.
  This means that struct rb_root must be compatible.  Add missing
  fields.  This fixes RB_REMOVE breakage due to it trying to iterate
  rbh_inprog.

* Add set_need_resched() (empty)

* radeon_bo_is_reserved() is no longer applicable and has been removed,
  allowing us to avoid implementintg ttm_bo_is_reserved().  Note that
  linux-current does not have these functions.

* Refactor radeon_do_test_moves() so reflect the linux code a bit
  better.  This fixes a few error paths.


    [22 lines not shown]

DragonFlyBSD/src 189951asys/dev/drm/ttm ttm_bo.c

drm/ttm: get rid of ttm_bo_is_reserved usage

DragonFlyBSD/src 6a00129sys/conf kern.pre.mk, sys/dev/drm/include/drm/ttm ttm_bo_driver.h ttm_object.h

drm/ttm: inline ttm_bo_reserve and related calls

DragonFlyBSD/src 94ff63esys/vm vnode_pager.c

sys/vm: Fix vnode_pager_haspage() description

Not uptodate with the code since
1b9d3514e66df9d4bc5a25c898c7c8cbe414e3e4 in 2010.

DragonFlyBSD/src 64d15basys/dev/misc/psm psm.c

psm - Fix panic in ps/2 mouse driver

* Fix a race in the ps/2 driver where a callout could be interrupted
  by psmintr() and corrupt the ps/2 packet buffer, causing a panic.

* Use a lockmgr lock instead of (archaic) critical sections for interrupt
  protection.  Also use the locked callout API.  This will hopefully
  prevent any further corruption.

Reported-by: drill-use at irc
DeltaFile
+49-48sys/dev/misc/psm/psm.c
+49-481 files

DragonFlyBSD/src 1ea2893sys/dev/misc/psm psm.c

psm - Fix panic in ps/2 mouse driver

* Fix a race in the ps/2 driver where a callout could be interrupted
  by psmintr() and corrupt the ps/2 packet buffer, causing a panic.

* Use a lockmgr lock instead of (archaic) critical sections for interrupt
  protection.  Also use the locked callout API.  This will hopefully
  prevent any further corruption.

Reported-by: drill-use at irc
DeltaFile
+49-48sys/dev/misc/psm/psm.c
+49-481 files

DragonFlyBSD/src d0975e2share/man/man9 idr.9, sys/libkern linux_idr.c

kernel - Fix IDR bugs

* Allow ptr to be NULL in idr_remove() and idr_replace().

* Note that linux ERR returns from idr_replace() still not implemented
  (requires the related ERR macros to be moved out of drm).  idr_replace()
  still returns NULL when the id cannot be found.

Requested-by: aly

DragonFlyBSD/src 3c3b358sys/dev/drm/include/drm/ttm ttm_bo_api.h ttm_bo_driver.h, sys/dev/drm/ttm ttm_bo.c ttm_execbuf_util.c

drm/ttm: convert to the reservation api

Additional-bugfix-by: dillon

DragonFlyBSD/src 6cd4d95sys/conf files, sys/dev/drm linux_wwmutex.c

drm: Fully rework the ww_mutex implementation

Fixing various bugs and inaccurate assertions

DragonFlyBSD/src 712774asys/vfs/ufs ufs_vnops.c dinode.h

ufs - Expand time_t support to 48 bits

* Fix time overflow issues in the original 32-bit UFS code in two
  ways.  First, treat the original 32-bit seconds fields as unsigned.
  Second, utilize the spare fields to expand these fields to 48 bits
  each.  Retain the nanosecond-grain accuracy of the nsec fields.

  These changes should be both forwards and backwards compatible within
  the respective supported timestamp ranges for filesystems created by
  DragonFlyBSD and for legacy UFS1 systems.

  However, DFly will no longer be compatible with any modern UFS1
  implementation in other operating systems which reuse the spare fields
  for other purposes.

  Also, dump and restore are still limited to 32-bit time stamps.  This
  should not be an issue since these utilities are basically no longer used
  in modern day.  UFS1 is still useful, but deprecated, and we don't really
  intend for it to be used with later incarnations that might still be in
  use by other OSes.  We do want to make sure that the time fields work
  properly, however, as UFS1 is still useful for certain things.

* The original di_atime, di_mtime, and di_ctime fields were implemented
  as signed 32 bit fields.  Change these to unsigned, extending time_t
  support from 2036 into the 2100 range.

    [6 lines not shown]

DragonFlyBSD/src 0749598gnu/usr.bin/groff/tmac mdoc.local

mdoc: Add FreeBSD 11.1 for efidev(4).

DragonFlyBSD/src 6be3395sys/conf files, sys/dev/drm/drm Makefile

drm: CONFIG_AGP is off by default