NetBSD/src xk8OlgI — sys/arch/m68k/m68k mmu_subr.s

   mmu_load_urp51(): We're switching the user address space by loading
   %crp, so only invalidate user ATC entries, not the entire ATC.
VersionDeltaFile
1.8+3-3sys/arch/m68k/m68k/mmu_subr.s
+3-31 files

NetBSD/src jVKNnTQ — sys/arch/m68k/m68k mmu_subr.s

   mmu_load_urp20hp(): No need to explicitly invalidate the external VAC
   or the MMU's TLB -- user VAC entries and user TLB entries are automatically
   invalidated when the USTP register is written, and there's no need to
   invalidate supervisor entries in either of those when switching the user
   address space.
VersionDeltaFile
1.7+4-9sys/arch/m68k/m68k/mmu_subr.s
+4-91 files

NetBSD/src DQdSzJC — sys/dev/acpi pckbc_acpi.c

   pckbc(4) at acpi: Fix build with -Os.

   Fixes:

   /home/riastradh/netbsd/current/src/sys/dev/acpi/pckbc_acpi.c: In function 'pckbc_acpi_intr_establish':
   /home/riastradh/netbsd/current/src/sys/dev/acpi/pckbc_acpi.c:242:63: error: 'handle' may be used uninitialized [-Werror=maybe-uninitialized]
     242 |                 rv = acpi_intr_establish(sc->sc_dv, (uint64_t)(uintptr_t)handle,
         |                                                               ^~~~~~~~~~~~~~~~~
   /home/riastradh/netbsd/current/src/sys/dev/acpi/pckbc_acpi.c:224:21: note: 'handle' was declared here
     224 |         ACPI_HANDLE handle;
         |                     ^~~~~~

   No functional change intended -- the code wasn't _wrong_ before; this
   just makes it clearer, with less analysis effort, that handle is
   initialized in the path in question.

   While here, rename local variable `rv' to `ih' to make it clearer
   what it is (not just some arbitrary return value of a subroutine but
   specifically an interrupt handler cookie).

    [3 lines not shown]
VersionDeltaFile
1.41+9-9sys/dev/acpi/pckbc_acpi.c
+9-91 files

NetBSD/src be2gn3U — usr.sbin/cpuctl/arch aarch64.c

   cpuctl: aarch64: name ID_AA64DFR0_EL1 correctly in a comment
VersionDeltaFile
1.32+3-3usr.sbin/cpuctl/arch/aarch64.c
+3-31 files

NetBSD/src gN2dLVN — usr.sbin/cpuctl/arch aarch64.c

   cpuctl: aarch64: fix id_aa64isar0.tlb field name
VersionDeltaFile
1.31+3-3usr.sbin/cpuctl/arch/aarch64.c
+3-31 files

NetBSD/src UFah2rP — usr.sbin/cpuctl/arch aarch64.c

   cpuctl: aarch64: put the string continuation character on the first line.

   NFCI
VersionDeltaFile
1.30+42-42usr.sbin/cpuctl/arch/aarch64.c
+42-421 files

NetBSD/src MBrlfiE — usr.sbin/cpuctl/arch aarch64.c

   cpuctl: aarch64: end all array element initialisers with ','

   NFCI
VersionDeltaFile
1.29+79-79usr.sbin/cpuctl/arch/aarch64.c
+79-791 files

NetBSD/src pedpHr0 — sys/arch/m68k/m68k pmap_68k.c

   Overhaul ATC / cache manipulation.

   ATC + cache invalidation information is now gathered in a per-CPU
   information struture that gathers up to 8 VAs in a batch.  If more
   than 8 VAs need to be invalidated, we just hit the entire ATC, trading
   off the cost of the per-entry vs all-entries operations.

   The same change allows us to batch cache invalidations, which reduces
   the number of redundant cache invalidations.

   Another nice property of these changes is that it collapses the HP MMU
   handling into the same code paths as garden-variety 68851/68030, simplifying
   the code flow in all sorts of places.
VersionDeltaFile
1.101+823-290sys/arch/m68k/m68k/pmap_68k.c
+823-2901 files

NetBSD/src kKbef3Y — lib/libc/stdlib strfmon.c

   pass lint
VersionDeltaFile
1.26+4-6lib/libc/stdlib/strfmon.c
+4-61 files

NetBSD/src jRk8WyF — usr.sbin/npf/npftest/libnpftest npf_ext_test.c

   use ret from raw test case in verbose to see return value

   and then tests can reach normalization tests for in flight packet testing.

   more in flight packets tests for routing and possibly logging soon
VersionDeltaFile
1.7+0-2usr.sbin/npf/npftest/libnpftest/npf_ext_test.c
+0-21 files

NetBSD/src S6C70WO — usr.sbin/npf/npftest/libnpftest npf_ext_test.c

   sprinkle CHECK_TRUE
VersionDeltaFile
1.6+3-5usr.sbin/npf/npftest/libnpftest/npf_ext_test.c
+3-51 files

NetBSD/src HWMI5Ap — sys/net/npf npf_ext_route.c

   set your flags before increasing stats
VersionDeltaFile
1.21+2-3sys/net/npf/npf_ext_route.c
+2-31 files

NetBSD/src B1ElmYI — usr.sbin/npf/npftest/libnpftest npf_bpf_test.c

   NPF bpf tests

   use array byte size for bpf_validate
VersionDeltaFile
1.16+1-1usr.sbin/npf/npftest/libnpftest/npf_bpf_test.c
+1-11 files

NetBSD/src Ce0xODd — usr.sbin/npf/npftest/libnpftest npf_nat_test.c

   use right port number in NAT test.

   mistakenly entered dst port instead of src after translation
VersionDeltaFile
1.16+1-1usr.sbin/npf/npftest/libnpftest/npf_nat_test.c
+1-11 files

NetBSD/src EabfqZN — external/gpl3/binutils/dist/gas/config tc-vax.c

   gas: vax: Cherry-pick upstream commit 2722595890a

   PR port-vax/59326
   ---
   Author: Jan-Benedict Glaw <jbglaw at getslash.de>
   Date:   Sun Nov 15 18:36:27 2026 +0100

   VAX: In PIC code, use GOT for function pointers

   Use GOT references for function pointers in PIC code. Actual fix by
   Rin Okuyama, initially reported for NetBSD in https://gnats.netbsd.org/59326 .

   PR gas/32941

   gas/
           * config/tc-vax.c (md_estimate_size_before_relax): Use GOT instead of PLT.
           * testsuite/gas/vax/elf-got-func.s,
           * testsuite/gas/vax/elf-got-func.d: New test.
           * testsuite/gas/vax/vax.exp: Run it.
VersionDeltaFile
1.25+8-3external/gpl3/binutils/dist/gas/config/tc-vax.c
+8-31 files

NetBSD/src fO2j1a9 — external/gpl3/binutils/dist/gas/config tc-vax.c

   gas: vax: Revert 1.22 in preparation to cherry-pick upstream fix

   Will be NFC (comment-only changes) in the end.

   PR port-vax/59326
VersionDeltaFile
1.24+1-4external/gpl3/binutils/dist/gas/config/tc-vax.c
+1-41 files

NetBSD/src kCtCE1g — usr.sbin/cpuctl/arch aarch64.c

   whitespace
VersionDeltaFile
1.28+36-36usr.sbin/cpuctl/arch/aarch64.c
+36-361 files

NetBSD/src T2S47fF — sys/arch/arm/sunxi sunxi_gpio.h sunxi_gpio.c, sys/arch/evbarm/conf GENERIC64

   sunxi: Add support for Allwinner A733 (sun60i-w2) PPU and pinctrl.

   Compile tested only at this point.
VersionDeltaFile
1.1+264-0sys/arch/arm/sunxi/sun60i_w2_gpio.c
1.1+168-0sys/arch/arm/sunxi/sun60i_w2_ppu.c
1.40+8-2sys/arch/arm/sunxi/sunxi_gpio.c
1.76+9-1sys/arch/arm/sunxi/files.sunxi
1.12+6-1sys/arch/arm/sunxi/sunxi_gpio.h
1.234+3-1sys/arch/evbarm/conf/GENERIC64
+458-56 files

NetBSD/src OYOa2yt — sys/arch/evbarm/ifpga ifpga_intr.h

   KNF
VersionDeltaFile
1.10+2-2sys/arch/evbarm/ifpga/ifpga_intr.h
+2-21 files

NetBSD/src rfUtgDz — share/man/man4 ds2413gpio.4

   ds2413gpio.4: fix a few typos
VersionDeltaFile
1.2+5-5share/man/man4/ds2413gpio.4
+5-51 files

NetBSD/src 3HELDNl — doc CHANGES


   Mention ds2413gpio(4)
VersionDeltaFile
1.3313+4-1doc/CHANGES
+4-11 files

NetBSD/src 3POLkH5 — share/man/man4 ds2413gpio.4, sys/dev/onewire files.onewire ds2413gpioreg.h


   A driver for the Maxim DS2413 1-Wire Dual Channel Addressable Switch,
   that is, two open drain GPIO ports at the end of a 1-Wire bus.

   Except for overdrive speed support, which is currently lacking in the
   onewire(4) infrastructure, all aspects of the chip are supported.  The
   ports are presented as a gpio(4) bus and are operated on like any GPIO
   bus would be with gpioctl.

   There is an official family ID for the DS2413, but widely known clones
   use an alternative.  Both IDs are supported.  It is also possible to
   add another ID using a kernel config option, as there are some reports
   of yet other family IDs out in the wild.

   The chip works on the DS248x I2C to 1-Wire bridges and works on the
   bitbanged gpioow driver.  Tests were performed on a Virtualbox amd64
   VM using umcpmio(4) as a USB to I2C bridge together with a DS2484 I2C
   to 1-Wire bridge as well as directly on a RPI2 and RPI3 running

    [7 lines not shown]
VersionDeltaFile
1.1+417-0sys/dev/onewire/ds2413gpio.c
1.1+78-0share/man/man4/ds2413gpio.4
1.1+44-0sys/dev/onewire/ds2413gpiovar.h
1.1+34-0sys/dev/onewire/ds2413gpioreg.h
1.1+15-0sys/modules/ds2413gpio/Makefile
1.7+7-1sys/dev/onewire/files.onewire
+595-18 files not shown
+622-1014 files

NetBSD/src PVIwgW9 — usr.sbin/npf/npfctl npf.conf.5

   npf.conf.5: terminate outer list correctly

   (Follow-on for PR misc/60946.)
VersionDeltaFile
1.116+2-1usr.sbin/npf/npfctl/npf.conf.5
+2-11 files

NetBSD/src 9HUHn3r — usr.bin/xlint/lint1 tree.c

   lint: fix integer constraints for integer ranges
VersionDeltaFile
1.720+4-4usr.bin/xlint/lint1/tree.c
+4-41 files

NetBSD/src ohmisg9 — usr.bin/xlint/lint1 tree.c

   lint: fix integer constraints for division

   In the expression 'a / b' with b > 0, lint wrongly assumed the result
   would have the same bits cleared as a.

   In the expression 'a % b', the resulting constraints were wrong for
   negative b.
VersionDeltaFile
1.719+31-18usr.bin/xlint/lint1/tree.c
+31-181 files

NetBSD/src 8foMDWy — sys/net/npf npf_impl.h npf.h

   Introduce consumed flags by procedures in NPF.

   Route procedures will almost always consume mbufs.
   introcude a flag in the nbuf to let us know
   that the nbuf has been consumed in case there is a
   NULL check.

   looked by martin
VersionDeltaFile
1.20+8-11sys/net/npf/npf_ext_route.c
1.58+8-6sys/net/npf/npf_handler.c
1.25+1-7sys/net/npf/npf_rproc.c
1.89+0-1sys/net/npf/npf_impl.h
1.71+1-0sys/net/npf/npf.h
+18-255 files

NetBSD/src czAc0gk — doc CHANGES-11.2

   Tickets #579 - #583
VersionDeltaFile
1.1.2.3+39-1doc/CHANGES-11.2
+39-11 files

NetBSD/src dM3r34A — sys/dev/ic wdc.c

   Pull up following revision(s) (requested by tls in ticket #583):

        sys/dev/ic/wdc.c: revision 1.318

   Fix typo in 1.313/1.311
VersionDeltaFile
1.310.12.3+3-3sys/dev/ic/wdc.c
+3-31 files

NetBSD/src eFR5vpC — share/man/man4 ds2482ow.4, sys/dev/i2c ds2482owreg.h ds2482owvar.h

   Pull up following revision(s) (requested by brad in ticket #582):

        sys/dev/i2c/ds2482owreg.h: revision 1.2
        sys/dev/onewire/onewire.c: revision 1.23
        share/man/man4/ds2482ow.4: revision 1.2
        sys/dev/i2c/ds2482owvar.h: revision 1.4
        sys/dev/i2c/ds2482ow.c: revision 1.3

   20 years ago when the 1-Wire infrastructure was ported over from
   OpenBSD, the calls in the onewire_detach() function in the onewirebus
   (sys/dev/onewire/onewire.c) code for config_detach_children() was
   commented out.  I don't know why this was done, exactly, as there
   wasn't any particular comments present.  I suspect it might have been
   because owtemp, the common end device, could not detach or maybe
   detaching just didn't work as expected at that time.

   The string of devices are as follows:
   End controller     -> onewirebus -> End device
   gpioow or ds2482ow ->            -> such as owtemp or ds24e17iic

    [28 lines not shown]
VersionDeltaFile
1.2.2.1+368-66sys/dev/i2c/ds2482ow.c
1.1.2.1+57-5share/man/man4/ds2482ow.4
1.3.2.1+19-10sys/dev/i2c/ds2482owvar.h
1.22.14.1+3-4sys/dev/onewire/onewire.c
1.1.2.1+3-1sys/dev/i2c/ds2482owreg.h
+450-865 files

NetBSD/src ZYfTmmi — share/mk bsd.own.mk

   Pull up following revision(s) (requested by riastradh in ticket #580):

        share/mk/bsd.own.mk: revision 1.1491

   hppa: Enable fortify/ssp in userland for progs/libs that opt into it.

   The `ssp' part doesn't actually do much (and really I'm not sure why
   these options are mixed together) because hppa stack grows up, so
   writing past the end of an array won't overwrite the return address,
   so stack smashing in this way is extremely unlikely.  But the
   `fortify' part, which instruments functions like memcpy/read/getcwd
   with bounds checks where statically known, is still useful to catch
   buffer overruns.

   ok `I guess' nick@

   PR lib/60858: fortuitous embarrassment: fortify is all kinds of busted
VersionDeltaFile
1.1428.2.6+1-2share/mk/bsd.own.mk
+1-21 files