NetBSD/src CVk4rPpsys/dev/pci pci_resource.c

   pci_resource: use existing BARs even if mem/io space is left disabled

   It's possible that firmware assigned resources to a device and left
   memory / io space accesses disabled. Let's assume if we see a non-zero
   address in a BAR that it was assigned by firmware.

   Should help PR# kern/60427 and kern/60463
VersionDeltaFile
1.8+11-2sys/dev/pci/pci_resource.c
+11-21 files

NetBSD/src f4eP8XQsys/arch/arm/cortex gicv3.c

   gicv3: don't wait for RWP to clear when unblocking irqs

   The RWP bit does not track writes to ISENABLER<n>, only ICENABLER<n>. So
   don't bother waiting for it to clear when unblocking IRQs.

   This is ok; pic_block_irqs is expected to be synchronous (no IRQs allowed
   after the call) but not pic_unblock_irqs (some time can pass before an IRQ
   is raised).
VersionDeltaFile
1.59+3-7sys/arch/arm/cortex/gicv3.c
+3-71 files

NetBSD/src 6eSW7jSdoc CHANGES-10.2

   Tickets #1338 and #1339
VersionDeltaFile
1.1.2.131+29-1doc/CHANGES-10.2
+29-11 files

NetBSD/src f6b1lIhexternal/public-domain/tz/dist asia southamerica

   Pull up the following, requested by kre in ticket #1339:

        external/public-domain/tz/dist/CONTRIBUTING     up to 1.1.1.11
        external/public-domain/tz/dist/Makefile         up to 1.7
        external/public-domain/tz/dist/NEWS             up to 1.9
        external/public-domain/tz/dist/README           up to 1.4
        external/public-domain/tz/dist/TZDATA_VERSION   up to 1.47
        external/public-domain/tz/dist/asia             up to 1.17
        external/public-domain/tz/dist/backward         up to 1.13
        external/public-domain/tz/dist/backzone         up to 1.5
        external/public-domain/tz/dist/northamerica     up to 1.8
        external/public-domain/tz/dist/southamerica     up to 1.6
        external/public-domain/tz/dist/theory.html      up to 1.7
        external/public-domain/tz/dist/version          up to 1.20
        external/public-domain/tz/dist/zone.tab         up to 1.7
        external/public-domain/tz/dist/zone1970.tab     up to 1.8
        external/public-domain/tz/dist/zonenow.tab      up to 1.9
        doc/3RDPARTY                                    (apply patch)

   Updated tzdata to 2026d (via local 2026dgtz).
VersionDeltaFile
1.1.1.32.2.8+80-46external/public-domain/tz/dist/northamerica
1.1.1.40.2.8+114-1external/public-domain/tz/dist/NEWS
1.1.1.17.2.6+32-31external/public-domain/tz/dist/theory.html
1.10.2.3+49-5external/public-domain/tz/dist/backward
1.1.1.22.2.7+16-13external/public-domain/tz/dist/southamerica
1.7.2.7+18-6external/public-domain/tz/dist/asia
+309-10210 files not shown
+336-12716 files

NetBSD/src 2kDukvcdoc CHANGES-11.1

   Tickets #454 - #464
VersionDeltaFile
1.1.2.12+86-1doc/CHANGES-11.1
+86-11 files

NetBSD/src Om2z0rysys/arch/xen/xen xbdback_xenbus.c

   Pull up following revision(s) (requested by bouyer in ticket #1338):

        sys/arch/xen/xen/xbdback_xenbus.c: revision 1.109

   Fix a brain-o in a KASSERT that could cause a guest to panic a DIAGNOSTIC
   kernel: to assert that an I/O is already in flight we want
   xbdi_pendingreqs > 0, not xbdi_pendingreqs > 1

   While there, fix an error path that could also cause a KASSERT to fire
   or a state corruption: reset xio_need_bounce to 0 before calling
   xbdback_io_error() when the bounce buffer is not fully set up.
VersionDeltaFile
1.101.4.3+4-3sys/arch/xen/xen/xbdback_xenbus.c
+4-31 files

NetBSD/src hcDRCNDsys/arch/xen/xen xbdback_xenbus.c

   Pull up following revision(s) (requested by bouyer in ticket #464):

        sys/arch/xen/xen/xbdback_xenbus.c: revision 1.109

   Fix a brain-o in a KASSERT that could cause a guest to panic a DIAGNOSTIC
   kernel: to assert that an I/O is already in flight we want
   xbdi_pendingreqs > 0, not xbdi_pendingreqs > 1

   While there, fix an error path that could also cause a KASSERT to fire
   or a state corruption: reset xio_need_bounce to 0 before calling
   xbdback_io_error() when the bounce buffer is not fully set up.
VersionDeltaFile
1.107.4.1+4-3sys/arch/xen/xen/xbdback_xenbus.c
+4-31 files

NetBSD/src w6EsCHF. UPDATING, lib/libc/time zic.c

   Apply patch, requested by kre in ticket #463:

        lib/libc/time/zic.c     (apply patch)
        UPDATING                (apply patch)

   Apply upstream fix for zic(8). PR 60715.
VersionDeltaFile
1.383.2.1+7-1UPDATING
1.95.2.1+3-3lib/libc/time/zic.c
+10-42 files

NetBSD/src ay6lCl9bin/stty print.c

   Pull up following revision(s) (requested by tls in ticket #462):

        bin/stty/print.c: revision 1.24

   Fix mis-printing of ONOCR and ONLRET.
VersionDeltaFile
1.23.40.1+4-4bin/stty/print.c
+4-41 files

NetBSD/src WOoL01esys/net bpf.c bpf_filter.c, sys/net/npf files.npf npf_os.c

   Pull up following revision(s) (requested by tls in ticket #461):

        sys/net/npf/files.npf: revision 1.26
        sys/net/bpf_filter.c: revision 1.74
        sys/net/bpf.c: revision 1.259
        sys/net/npf/npf_os.c: revision 1.24
        sys/net/npf/npf_os.c: revision 1.25

   Fix npf crash in kernels without "pseudo-device bpfilter".

   Actually fix panic on non-modular kernels with npf and no bpf.

   Fix commit of badly-applied patch that brought us npf_os.c 1.24.
VersionDeltaFile
1.73.2.1+36-2sys/net/bpf_filter.c
1.258.2.1+2-28sys/net/bpf.c
1.23.2.1+3-3sys/net/npf/npf_os.c
1.24.2.1+2-2sys/net/npf/files.npf
+43-354 files

NetBSD/src eo293Jasys/net/npf npf_ext_route.c

   npf_ext_route.c: s/fragmenrt/fragment in comment
VersionDeltaFile
1.7+2-2sys/net/npf/npf_ext_route.c
+2-21 files

NetBSD/src wUiFFlqsys/net/npf npf_ext_route.c

   - only zero the nbuf if we counsumed the mbuf
   - remove KASSERT (it was only the the v4 case)
   - fix indentation
   - merge success/failure paths
VersionDeltaFile
1.6+22-21sys/net/npf/npf_ext_route.c
+22-211 files

NetBSD/src EfuKPYFsys/arch/aarch64/aarch64 fpu.c

   arm64: fpu: avoid unnecessarily reads of MVFR1_EL1

   The contents of MVFR1_EL1 do not change at runtime and it is already
   saved in struct cpu_info. Use the cached value in fpu_state_load() as it
   avoids a guest exit when running under a hypervisor with HCR_EL2.TID3 = '1'.
VersionDeltaFile
1.14+3-3sys/arch/aarch64/aarch64/fpu.c
+3-31 files

NetBSD/src muOQz4rsys/dev/i2c lm75.c

   Pull up following revision(s) (requested by jdc in ticket #460):

        sys/dev/i2c/lm75.c: revision 1.51

   Fix setting limits:
     check for NULL limits pointer (restore)
     handle degc (user) or raw (restore initial) values
VersionDeltaFile
1.47.2.1+33-19sys/dev/i2c/lm75.c
+33-191 files

NetBSD/src cARJQM4share/man/man4 adm1026hm.4, sys/dev/i2c adm1026reg.h adm1026.c

   Pull up following revision(s) (requested by jdc in ticket #459):

        sys/dev/i2c/adm1026.c: revision 1.14
        sys/dev/i2c/adm1026.c: revision 1.15
        sys/dev/i2c/adm1026.c: revision 1.16
        sys/dev/i2c/adm1026.c: revision 1.17
        sys/dev/i2c/adm1026reg.h: revision 1.6
        share/man/man4/adm1026hm.4: revision 1.2
        share/man/man4/adm1026hm.4: revision 1.3
        share/man/man4/adm1026hm.4: revision 1.4

   Remove the "multi_read" logic now that the problem in the pcf8584 driver
   has been fixed (r1.23 - use repeat start).

   Handle cases where we have a gap in the sensors configured/present with
   a map of envsys sensor to chip sensor.

   Use a variable to hold the register number instead of passing a constant
   to the iic routines.

    [26 lines not shown]
VersionDeltaFile
1.13.18.1+573-221sys/dev/i2c/adm1026.c
1.5.28.1+31-31sys/dev/i2c/adm1026reg.h
1.1.30.1+37-19share/man/man4/adm1026hm.4
+641-2713 files

NetBSD/src qeHzcghsys/arch/sparc64/sparc64 ofw_patch.h autoconf.c

   Pull up following revision(s) (requested by jdc in ticket #458):

        sys/arch/sparc64/sparc64/autoconf.c: revision 1.247 (patch)
        sys/arch/sparc64/sparc64/ofw_patch.c: revision 1.10 (patch)
        sys/arch/sparc64/sparc64/ofw_patch.h: revision 1.5

   Reduce the amount of patching conditions in autoconf.c by moving the
   machine model checks to ofw_patch.c.  We now have one patch routine for
   i2c buses and another for i2c devices.

   Consolidate the patching checks for parent bus type into a single check.

   Remove the "multi_read" property for V210, V240 and V440 - no longer needed
   because the pcfiic driver was fixed.

   While here, add a "fan_mask" property for the V240, which has only 3 fans
   in the first group of 4.
VersionDeltaFile
1.7.26.1+48-25sys/arch/sparc64/sparc64/ofw_patch.c
1.243.2.1+5-32sys/arch/sparc64/sparc64/autoconf.c
1.4.28.1+3-8sys/arch/sparc64/sparc64/ofw_patch.h
+56-653 files

NetBSD/src xkPFobIsys/arch/sparc64/conf files.sparc64 GENERIC, sys/arch/sparc64/dev bq4802_ebus.c bq4802_ebus.c

   Pull up following revision(s) (requested by jdc in ticket #457):

        sys/dev/ic/bq4802reg.h: revision 1.1
        sys/dev/ic/bq4802reg.h: revision 1.2
        sys/arch/sparc64/conf/GENERIC: revision 1.245 (via patch)
        sys/arch/sparc64/conf/files.sparc64: revision 1.169 (via patch)
        sys/arch/sparc64/dev/bq4802_ebus.c: revision 1.1
        sys/arch/sparc64/dev/bq4802_ebus.c: revision 1.2

   Add a driver for the Texas Instruments bq4802Y/bq4802LY RTC chip,
   used in the Sun Ultra 45.

   Add bq4802rtc for the Ultra 45 RTC.

   Add bq4802rtc (RTC for U45).

   Use the correct setting for the stop bit, so that the clock runs when the
   power is off (instead of vice versa).  Make this clearer in the header.
VersionDeltaFile
1.3.2.1+0-265sys/arch/sparc64/dev/bq4802_ebus.c
1.3.2.2+184-0sys/arch/sparc64/dev/bq4802_ebus.c
1.3.2.1+0-175sys/dev/ic/bq4802reg.h
1.3.2.2+175-0sys/dev/ic/bq4802reg.h
1.243.2.2+5-2sys/arch/sparc64/conf/GENERIC
1.166.20.1+5-1sys/arch/sparc64/conf/files.sparc64
+369-4436 files

NetBSD/src lxCpHp3sys/arch/sparc64/dev pcfiic_ebus.c, sys/dev/ic pcf8584var.h pcf8584.c

   Pull up following revision(s) (requested by jdc in ticket #456):

        sys/dev/ic/pcf8584var.h: revision 1.7 (patch)
        sys/arch/sparc64/dev/pcfiic_ebus.c: revision 1.10 (patch)
        sys/dev/ic/pcf8584.c: revision 1.23 (patch)

   When reading from an i2c device register, use repeat start on the bus.

   This avoids the situation where another master gets the bus between our
   register select (i.e. write) and the following read.

   Add a delay after accessing our registers for the PIC implementation
   of the PCF8584 (e.g. on the Sun Ultra 45).

   Change swapregs to flags to handle multiple quirks and set the flags
   appropriately in the Ebus frontend.
VersionDeltaFile
1.20.12.1+67-24sys/dev/ic/pcf8584.c
1.7.28.1+9-5sys/arch/sparc64/dev/pcfiic_ebus.c
1.6.34.1+6-1sys/dev/ic/pcf8584var.h
+82-303 files

NetBSD/src yfJR2oUsys/uvm uvm_pdpolicy_clock.c uvm_pdpolicy_clockpro.c

   Pull up following revision(s) (requested by tls in ticket #455):

        sys/uvm/uvm_pdpolicy_clockpro.c: revision 1.28
        sys/uvm/uvm_pdpolicy_clockpro.c: revision 1.29
        sys/uvm/uvm_pdpolicy_clockpro.c: revision 1.30
        sys/uvm/uvm_pdpolicy_clock.c: revision 1.43

   Fix a very brief race which could cause page states to be corrupted.

   Observed under heavy physio load from an 8-way parallel dump|restore
   workload used in filesystem testing - manifested as intents being
   scribbled over while the interlock was momentarily not held
   in uvmpdpol_selectvictim.  I've attempted to apply the same fix to
   clockpro, which looks like it needs it too, but clockpro was on fire
   when I got there and I didn't put it out (still does not compile).

   Fix an obvious typo which applied the wrong intent.  Still does not compile.

   fix build with clockpro
VersionDeltaFile
1.27.12.1+56-12sys/uvm/uvm_pdpolicy_clockpro.c
1.42.2.1+32-3sys/uvm/uvm_pdpolicy_clock.c
+88-152 files

NetBSD/src gDdWMCNsys/compat/linux/common linux_file.c

   Pull up following revision(s) (requested by sborrill in ticket #454):

        sys/compat/linux/common/linux_file.c: revision 1.135

   Fix edge-case panic in copy_file_range when off_in or off_out are not NULL.

   Found by Henrique Brito as part of GSoC 2024.
VersionDeltaFile
1.133.2.1+4-6sys/compat/linux/common/linux_file.c
+4-61 files

NetBSD/src It4j5Xgsys/arch/ofppc/ofppc machdep.c

   provide cmap_callback
VersionDeltaFile
1.123+22-2sys/arch/ofppc/ofppc/machdep.c
+22-21 files

NetBSD/src SZoo7pDsys/net/npf npf_ext_route.c

   PR kern/60721

   avoid icmp errors in routing path as icmp6_error also frees.

   this avoids double freeing
VersionDeltaFile
1.5+15-17sys/net/npf/npf_ext_route.c
+15-171 files

NetBSD/src VAbbcPCsys/net/npf npf_ext_route.c

   PR kern/60721

   Avoid double freeing of outgoing packet when routing takes over.
   memset it before ip_output tries to free it again its final path.
VersionDeltaFile
1.4+3-5sys/net/npf/npf_ext_route.c
+3-51 files

NetBSD/src WnMI4Dtsys/net/npf npf_ext_route.c

   - Don't use both m0 and m when m would suffice.
   - Fix indent in a comment
   - Remove useless setting to NULL
VersionDeltaFile
1.3+18-22sys/net/npf/npf_ext_route.c
+18-221 files

NetBSD/src o3Ot3nksys/arch/sparc64/dev vpcivar.h vpci.c

   sun4v: vpci(4) should use hypervisor interrupt api wrapper functions.

   From OpenBSD.
VersionDeltaFile
1.17+51-25sys/arch/sparc64/dev/vpci.c
1.2+2-1sys/arch/sparc64/dev/vpcivar.h
+53-262 files

NetBSD/src dxG6Fzvsys/arch/sparc64/sparc64 autoconf.c

   sun4v: Use hypervisor interrupt api version 3.

   From OpenBSD.
VersionDeltaFile
1.255+8-14sys/arch/sparc64/sparc64/autoconf.c
+8-141 files

NetBSD/src DAvNbeBsys/arch/sparc64/sparc64 intr.c

   sun4v: Use the deprecated hypervisor interrupt api if major is less than 3.

   From OpenBSD.
VersionDeltaFile
1.75+7-7sys/arch/sparc64/sparc64/intr.c
+7-71 files

NetBSD/src kwhLTmHdoc CHANGES 3RDPARTY

   new tzcode
VersionDeltaFile
1.2247+2-2doc/3RDPARTY
1.3292+2-1doc/CHANGES
+4-32 files

NetBSD/src wrtTPeelib/libc/time zic.8 theory.html

   Patch to 2026d (previous was 2026c)

     Changes to code

       zic now rejects Link targets that would have invalid names, and
       more efficiently processes Expires, Leap and Rule lines with years
       far in the past or future.  (Thanks to Darren Carreras.)

       zic now ports to systems that report lack of link support via
       EINVAL, ENOSYS or EPERM errno values. (Thanks to Tom Lane.)

       When tzset and related functions encounter a TZif file that is too
       large for them, they now consistently fail instead of sometimes
       silently ignoring excess parts of the file.

       localtime-related functions no longer mishandle extreme timestamps
       when given TZif files holding some unlikely timezone histories.
       (Problem reported by David Sarkisyan.)


    [12 lines not shown]
VersionDeltaFile
1.156+177-244lib/libc/time/localtime.c
1.24+60-58lib/libc/time/tz-link.html
1.52+114-1lib/libc/time/NEWS
1.102+54-45lib/libc/time/zic.c
1.25+32-31lib/libc/time/theory.html
1.56+29-15lib/libc/time/zic.8
+466-3949 files not shown
+526-42915 files

NetBSD/src cl9yJqUsys/arch/aarch64/aarch64 vmt.c, sys/dev/acpi acpi_util.h acpi_util.c

   PR/60655: vmt(4) probes the VMware backdoor unconditionally on aarch64

   On ACPI systems only probe for vmt(4) when Hypervisor_Id in FADT reports
   VMWare.

   This helps avoid buggy hypervisors that don't implement a RAZ mdccsr_el0

   Patch from Showta Ishizaki and slightly updated by me.
VersionDeltaFile
1.38+41-2sys/dev/acpi/acpi_util.c
1.3+26-1sys/arch/aarch64/aarch64/vmt.c
1.17+3-1sys/dev/acpi/acpi_util.h
+70-43 files