NetBSD/src NusWU7aexternal/mit/ctwm/libexec ctwm_app_menu

   Pull up following revision(s) (requested by nia in ticket #1253):

        external/mit/ctwm/libexec/ctwm_app_menu: revision 1.7

   ctwm_app_menu: Rewrite largely in awk for ~20-100x speedup.

   Previously took ~2.5sec on my laptop, now 0.03sec.
   Previously took ~10sec on a wiiu, now ~0.5sec.

   Output is meant to be byte-for-byte identical, except possibly in
   cases that could have screwed up ctwm by quoting shenanigans which
   are now escaped.  (I hope the escape sequences work, didn't actually
   check how ctwm interprets them.)  Can maybe support Exec line with
   `"' in them by deleting some code (marked XXX) but I didn't test that
   it actually works that way.

   PR bin/59958: ctwm: long delay during ctwm_app_menu
VersionDeltaFile
1.6.2.1+117-121external/mit/ctwm/libexec/ctwm_app_menu
+117-1211 files

NetBSD/src 8KksaH7common/lib/libc/arch/m68k/atomic atomic_init_m68k.c

   Remove the __sync_val_compare_and_swap_[12] aliases from m68k and use
   the new wrapper functions in

       common/lib/libc/atomic/atomic_cas_{8,16}_cas.c

   as all other that need these funcitions platforms do.
VersionDeltaFile
1.3+2-4common/lib/libc/arch/m68k/atomic/atomic_init_m68k.c
+2-41 files

NetBSD/src T7iw34Xexternal/mit/ctwm/libexec ctwm_app_menu

   Pull up following revision(s) (requested by nia in ticket #256):

        external/mit/ctwm/libexec/ctwm_app_menu: revision 1.7

   ctwm_app_menu: Rewrite largely in awk for ~20-100x speedup.

   Previously took ~2.5sec on my laptop, now 0.03sec.
   Previously took ~10sec on a wiiu, now ~0.5sec.

   Output is meant to be byte-for-byte identical, except possibly in
   cases that could have screwed up ctwm by quoting shenanigans which
   are now escaped.  (I hope the escape sequences work, didn't actually
   check how ctwm interprets them.)  Can maybe support Exec line with
   `"' in them by deleting some code (marked XXX) but I didn't test that
   it actually works that way.

   PR bin/59958: ctwm: long delay during ctwm_app_menu
VersionDeltaFile
1.6.6.1+117-121external/mit/ctwm/libexec/ctwm_app_menu
+117-1211 files

NetBSD/src 2OpSArxdistrib/notes/evbarm install

   Pull up following revision(s) (requested by nia in ticket #255):

        distrib/notes/evbarm/install: revision 1.19

   evbarm: Update install instructions.
   Recommend using installboot to install u-boot.
VersionDeltaFile
1.18.6.1+14-26distrib/notes/evbarm/install
+14-261 files

NetBSD/src YjRKys5libexec/ld.elf_so map_object.c rtld.h, libexec/ld.elf_so/arch/powerpc fixup.c fixup.c

   Pull up following revision(s) (requested by jmcneill in ticket #254):

        libexec/ld.elf_so/map_object.c: revision 1.70
        libexec/ld.elf_so/arch/powerpc/fixup.c: revision 1.1
        libexec/ld.elf_so/arch/powerpc/fixup.c: revision 1.2
        libexec/ld.elf_so/rtld.h: revision 1.152
        libexec/ld.elf_so/arch/powerpc/Makefile.inc: revision 1.18
        libexec/ld.elf_so/Makefile: revision 1.154

   This change adds the userspace portion of the IBM Espresso SMP fixup to the
   netbsd-11 branch. This enables SMP-safeness of a netbsd-11 userland running
   under a -current kernel on Wii U systems. Useful for eg. bulk builds in
   chroots.
VersionDeltaFile
1.3.2.1+0-201libexec/ld.elf_so/arch/powerpc/fixup.c
1.3.2.2+201-0libexec/ld.elf_so/arch/powerpc/fixup.c
1.69.2.1+9-2libexec/ld.elf_so/map_object.c
1.150.2.2+5-1libexec/ld.elf_so/rtld.h
1.151.2.2+4-1libexec/ld.elf_so/Makefile
1.17.4.1+2-2libexec/ld.elf_so/arch/powerpc/Makefile.inc
+221-2076 files

NetBSD/src ChzpzWobin/df df.c

   Pull up following revision(s) (requested by kim in ticket #253):

        bin/df/df.c: revision 1.106
        bin/df/df.c: revision 1.107

   df: Add human-readable inode counts

   df: Rename nbytes to value in prthumanval (not always bytes)
VersionDeltaFile
1.104.2.2+35-19bin/df/df.c
+35-191 files

NetBSD/src czY1ciRexternal/mpl/bind/dist/lib/isc jemalloc_shim.h

   Pull up following revision(s) (requested by tsutsui in ticket #251):

        external/mpl/bind/dist/lib/isc/jemalloc_shim.h: revision 1.6

   PR/60150: Izumi Tsutsui: Increase alignment so that libuv can use the bottom
   3 bits.
VersionDeltaFile
1.4.2.1+3-2external/mpl/bind/dist/lib/isc/jemalloc_shim.h
+3-21 files

NetBSD/src 2Vlrpbjsys/arch/m68k/m68k pmap_68k.c

   Add a comment describing why we seed the U and M bits in the PTE based
   on the access that resulted in the page fault that got us to pmap_enter().
VersionDeltaFile
1.62+8-2sys/arch/m68k/m68k/pmap_68k.c
+8-21 files

NetBSD/src mbSaGQisys/arch/m68k/m68k pmap_68k.c

   Update comments reflecting current status on mac68k.
VersionDeltaFile
1.61+8-6sys/arch/m68k/m68k/pmap_68k.c
+8-61 files

NetBSD/src CZ4V9ozsys/arch/mac68k/mac68k machdep.c macrom.c

   In the __HAVE_NEW_PMAP_68K case, use machine_bootmap[] to map the I/O
   area VA==PA.  With the Utah pmap, the I/O area was mapped at some
   arbitrary KVA in pmap_bootstrap() and then double-mapped VA==PA in
   mrg_init().  Initially, the new pmap case did this, as well, but then
   mrg_init() blew up because no page tables were present for the address
   it attempted to use.  Now that pmap_bootstrap1() ensures that PTs are
   allocated for FIXEDVA entries in machine_bootmap[], we can just have
   the one mapping.

   Confirmed by nat@ that this makes mac68k boot and run with the new
   pmap, but there are still gremlins when running the X server.
VersionDeltaFile
1.388+9-4sys/arch/mac68k/mac68k/machdep.c
1.77+8-2sys/arch/mac68k/mac68k/macrom.c
+17-62 files

NetBSD/src DJtC4A8libexec/httpd tilde-luzah-bozo.c

   Fix double free of hr_file (later handled by bozo_clean_request())

   Also hr_file_free should be used for free(3) (it fixes "//~user/" case
   in which hr_file points to the last slash in the prefix).
VersionDeltaFile
1.17+3-3libexec/httpd/tilde-luzah-bozo.c
+3-31 files

NetBSD/src a0tgx2qcommon/lib/libc/atomic atomic_init_testset.c atomic_cas_16_cas.c

   Move two functions from atomic_init_testset.c

   sync_val_compare_and_swap_1 --> atomic_cas_8_cas.c
   sync_val_compare_and_swap_2 --> atomic_cas_16_cas.c

   this helps with pullups
VersionDeltaFile
1.23+2-26common/lib/libc/atomic/atomic_init_testset.c
1.5+13-1common/lib/libc/atomic/atomic_cas_16_cas.c
1.6+13-1common/lib/libc/atomic/atomic_cas_8_cas.c
+28-283 files

NetBSD/src ZTuIMxVlib/libpthread pthread_types.h

   libpthread: comment unused fields unused
VersionDeltaFile
1.28+7-10lib/libpthread/pthread_types.h
+7-101 files

NetBSD/src We2YaUxlib/libpthread pthread_cond.c

   libpthread: replace &pthread__cond_dummy with an artificial invalid pointer

   a motivation is to make unexpected accesses trap.

   this would also save a bit of memory and relocation. (at least on x86)
VersionDeltaFile
1.79+7-7lib/libpthread/pthread_cond.c
+7-71 files

NetBSD/src 6meYKOAlib/libpthread pthread_mutex.c

   libpthread: retire struct waiter

   two identical structures, struct waiter and struct pthread__waiter,
   were introduced by a single commit in 2020.
   ("- Make pthread_condvar and pthread_mutex work on the stack...")
   since then, they have always been assumed to be identical.
   let's retire one of them and use a single copy.

   no functional changes are intended.
VersionDeltaFile
1.84+3-8lib/libpthread/pthread_mutex.c
+3-81 files

NetBSD/src FmTh4NMsys/arch/riscv/include ptrace.h vmparam.h

   Spell RISC-V correctly in comments.
VersionDeltaFile
1.7+2-2sys/arch/riscv/include/ptrace.h
1.15+2-2sys/arch/riscv/include/vmparam.h
1.8+2-2sys/arch/riscv/include/intr.h
1.2+2-2sys/arch/riscv/include/lwp_private.h
1.2+2-2sys/arch/riscv/include/pci_machdep.h
1.30+2-2sys/arch/riscv/include/pmap.h
+12-121 files not shown
+14-147 files

NetBSD/src cCpqyZmsys/arch/m68k/include pmap_68k.h pmap_motorola.h, sys/arch/m68k/m68k pmap_68k.c bus_dma.c

   NBPG -> PAGE_SIZE, PGSHIFT -> PAGE_SHIFT, PGOFSET -> PAGE_MASK
VersionDeltaFile
1.17+11-11sys/arch/m68k/include/pmap_68k.h
1.60+14-8sys/arch/m68k/m68k/pmap_68k.c
1.62+8-12sys/arch/m68k/include/pmap_motorola.h
1.44+7-7sys/arch/m68k/m68k/bus_dma.c
1.108+9-3sys/arch/m68k/m68k/pmap_motorola.c
1.19+5-5sys/arch/m68k/m68k/sys_machdep.c
+54-464 files not shown
+72-6410 files

NetBSD/src jquD50eusr.sbin/makefs makefs.8 TODO

   makefs(8): improve fs-options documentation

   Ensure all fs-options are documented, in alphabetical order.
   Describe whether the option has a value (e.g., foo=str or foo=num),
   and for numeric values describe the supported range.
   Some markup consistency.
   Remove my email from the man page.
VersionDeltaFile
1.76+202-142usr.sbin/makefs/makefs.8
1.8+1-5usr.sbin/makefs/TODO
+203-1472 files

NetBSD/src IR3UwXadistrib/utils/zcat Makefile

   define ZLIB_CONST so that this builds with new zlib.

   fixes sun2 build.
VersionDeltaFile
1.22+2-2distrib/utils/zcat/Makefile
+2-21 files

NetBSD/src xVtoQYOsys/lib/libkern/arch/m68k Makefile.inc, sys/lib/libkern/arch/mips Makefile.inc

   revert the previous, it broke all the m68k, mips, and powerpc builds.

   i had fixed most of the broken ones already, and i think thorpej has
   fixed the sun2/sun3 builds that were remaining broken.
VersionDeltaFile
1.26+2-2sys/lib/libkern/arch/mips/Makefile.inc
1.36+2-2sys/lib/libkern/arch/powerpc/Makefile.inc
1.40+1-2sys/lib/libkern/arch/m68k/Makefile.inc
+5-63 files

NetBSD/src Wp89bOYsys/lib/libkern/arch/m68k Makefile.inc, sys/lib/libkern/arch/mips Makefile.inc

   Add moddi3.c needed be new zlib adler32.c
VersionDeltaFile
1.25+2-2sys/lib/libkern/arch/mips/Makefile.inc
1.35+2-2sys/lib/libkern/arch/powerpc/Makefile.inc
1.39+2-1sys/lib/libkern/arch/m68k/Makefile.inc
+6-53 files

NetBSD/src FCZ28FAsys/arch/sun2/include pmap.h, sys/arch/sun3/include pmap3.h pmap3x.h

   Expose some declarations for _STANDALONE, as well.
VersionDeltaFile
1.31+16-13sys/arch/sun2/include/pmap.h
1.54+11-4sys/arch/sun3/include/pmap3.h
1.35+11-4sys/arch/sun3/include/pmap3x.h
+38-213 files

NetBSD/src JcGBAGesys/arch/virt68k/include param.h

   Fix a tyop in a comment.
VersionDeltaFile
1.5+2-2sys/arch/virt68k/include/param.h
+2-21 files

NetBSD/src XM4Slbtlib/libpthread pthread.c

   pthread__getstack: remove a redundant assignment

   no functional changes are intended.
VersionDeltaFile
1.196+0-2lib/libpthread/pthread.c
+0-21 files

NetBSD/src TZ42WBnlib/libpthread pthread.c

   libpthread: remove a stale comment

   the trick described by the comment has been removed in 2007.
   ("- Maintain a per-thread pointer to the last mutex acquired by...")
VersionDeltaFile
1.195+0-7lib/libpthread/pthread.c
+0-71 files

NetBSD/src CdoxgNSsys/arch/bebox/stand/boot Makefile, sys/arch/news68k/stand/boot Makefile

   fix several builds by making the library order libsa, libz, libkern.
VersionDeltaFile
1.38+2-2sys/arch/bebox/stand/boot/Makefile
1.28+2-2sys/arch/news68k/stand/boot/Makefile
1.34+2-2sys/arch/prep/stand/boot/Makefile
1.9+2-2sys/arch/rs6000/stand/boot/Makefile
1.35+2-2sys/arch/x68k/stand/boot/Makefile
1.32+2-2sys/arch/x68k/stand/libsa/Makefile
+12-126 files

NetBSD/src v6v6V2Nsys/net if.c if.h

   if: replace link state change queue with state transition

   if_link_queue is now a normal bitmask rather than holding a queue.
   It holds three bits to mirror the link state - UNKNOWN, DOWN and UP.
   There are also some bits to indicate that the link state has been scheduled
   for change and if it has been locked for changes (ie the interface being
   destroyed).

   The logic is simple - transitioning to DOWN will remove UNKNOWN and UP,
   transitioning to UNKNOWN will remove UP (no driver should do this).
   This means that even in the event of transitions happening faster than
   the kernel can spit them out, the correct state of the link will be
   preserved which is more important than the exact chain of events.

   This also fixes an issue where the workqueue for the link state change
   was incorrectly scheduled.

   if_link_scheduled is now unused and will be removed in a future patch.

   Fixes PR kern/60056.
VersionDeltaFile
1.537+66-113sys/net/if.c
1.310+7-2sys/net/if.h
+73-1152 files

NetBSD/src FUhSm6qlib/libpthread shlib_version pthread.c

   libpthread: explain why pthread__allqueue has been kept

   a bit history:

   - libpthread_dbg has been removed from base in 2017.

   - pkgsrc/devel/libpthread_dbg has been removed from pkgsrc in 2020.
     the commit message of the removal was calling it "Legacy library out
     of sync with NetBSD libpthread and without any users".
VersionDeltaFile
1.26+2-2lib/libpthread/shlib_version
1.194+1-1lib/libpthread/pthread.c
+3-32 files

NetBSD/src uakwxHNlib/libpthread pthread.c pthread_int.h

   libpthread: fix pthread_main_np

   the previous implementation was incorrect as pthread_t
   of the main thread can be recycled via pthread__deadqueue.

   with this changes, my test program [1] produces the same
   result as macOS.

   [1] https://github.com/yamt/garbage/tree/a24535e948dc62dde4c67366db97377e53d92e7b/c/pthread_main_np

   an alternative fix would be to stop recycling the main thread
   by simply leaking it. the leak is ok because, after all, it's
   very rare for applications to continue after its main thread
   exits. such applications don't even work as intended on some
   of pthread implementations in the wild. for now, i'd go with
   a more straightforward implementation though. ie. this commit.
VersionDeltaFile
1.193+9-7lib/libpthread/pthread.c
1.115+2-1lib/libpthread/pthread_int.h
+11-82 files

NetBSD/src ydzxCYFsys/fs/cd9660 cd9660_rrip.c

   cd9660: make sure that NM records are at least 5 bytes long.

   avoids an integer underflow when this length has 5 subtracted from it
   for a later path.

   Reported by Adam Crosser, Praetorian
VersionDeltaFile
1.19+8-2sys/fs/cd9660/cd9660_rrip.c
+8-21 files