Pull up following revision(s) (requested by kre in ticket #1330):
usr.bin/xinstall/xinstall.c: revision 1.131 (patch)
PR bin/58577 - install(1) -d issues
Fix issues where "install -d" (with no directory) simply
exit(0)s. That one is kind of marginal, installing nothing
when nothing is needed could be treated as OK, but the man
page does indicate in the SYNOPSIS that with -d, at least
one directory is needed (it says nothing at all about that
in the text).
Second, after creating a directory, if a later operation
(chown, chmod) fails, that is not success, a warning was
issued (good), a bad metalog was being created (bad).
That is clearly a bug (though probably doesn't happen
very often).
[29 lines not shown]
Pull up following revision(s) (requested by kre in ticket #1331):
lib/libc/locale/runetable.c: revision 1.30
lib/libc/locale/iswctype_mb.c: revision 1.15
PR lib/59067 (wctrans got error member)
Patches from the OP (ru_j217) and from RVP - see the PR
This looks to be just correcting what appear to be simple
errors in the code.
Pull up following revision(s) (requested by kre in ticket #1332):
external/mit/xorg/lib/driver.old.mk: revision 1.3
external/mit/xorg/lib/driver.mk: revision 1.10 (patch)
PR xsrc/59858 (locale fixes for xsrc)
From RVP - see the PR
Oversimplifying: this causes Mesa to use LC_NUMERIC=C when using
strtod() to parse stuff (ie: the radix char (decimal point) is '.',
regardless of the user's locale).
nfs: On reflection, just clamp the export_args buffer to 64 KiB.
Keep it simple, just make sure userland doesn't trigger contiguous
kva allocation of more than 64 KiB.
PR kern/60645: nfssvc(2): arithmetic overflow in input array sizing
mount(8) - add -W option (like df -W)
Add a -W option to mount, to show wedge names as labels (NAME=xyzzy)
instead of /dev/dkN (and if given twice, both forms), similar to the
-W option in df(1).
nfs: Work around missing VM_MIN/MAX_KERNEL_ADDRESS in rump.
Let's get the builds unbroken before making this perfect.
PR kern/60645: nfssvc(2): arithmetic overflow in input array sizing
Kernels ALL/i386 and ALL/amd64 warn "'ticket' may be used uninitialized"
at line 244, uvmfault_amapcopy(): "uvm_wait("fltamapcopy", ticket);".
Restructure to reflect the logic better and keep GCC quiet.
Ok: Taylor R Campbell
PR kern/58964: uvm: missing wakeup on uvmexp.free
PR kern/60029: panic: cpu0: softints stuck for 16 seconds
nfs: New sysctl node vfs.nfsd.maxexportspermount.
Limits the number of different ways a single mount point can be
exported. Default is limited to 256; can be expanded by sysctl up to
anything that won't overflow kva.
XXX Should prevent increasing it at securelevel>0.
PR kern/60645: nfssvc(2): arithmetic overflow in input array sizing
cpuctl(8): decode L2 cache associativity for VIA CPUs using the AMD cache lookup
table for the 0x80000006 CPUID function associativity value.
Use `cpuid leaf 4' to get cache parameters for Zhaoxin and late VIA CPU models.
Mirrors the changes from identcpu.c rev. 1.142.
Additionally, use via_cpu_probe() and via_cpu_cacheinfo() for CPUVENDOR_IDT
family > 6 CPUs to identify Zhaoxin CPUs and their features.
Also, print feature 5-7 bits for CPUVENDOR_IDT, which appear to be compatible
with Intel.
Use vendorname = "Zhaoxin" for family 7 models.
Tested on various 32-bit and 64-bit VIA and Zhaoxin CPUs
(Samuel 2 / Nehemiah, Esther (32-bit), multiple Nano models, and KX-U6580/7000).
PR kern/60619
pullups needed to netbsd-11 and netbsd-10.
Decode L2 cache associativity for VIA CPUs using the AMD cache lookup table for
the 0x80000006 CPUID function associativity value.
Use `cpuid leaf 4' to get cache parameters for Zhaoxin and late VIA CPU models.
VIA datasheets mention that associativity is stored as a decimal value. However,
in practice, it is encoded the AMD way and does not store the actual value.
Using the raw value leads to an incorrect CPU colors calculation, causing a
panic on KX-7000 (since the result is not a power of two) and incorrect CPU
colors values on other CPU models (e.g. as visible in `vmstat -s').
Decoding value using the AMD cache lookup table produces correct values in all
cases I have tested or verified using CPUID dumps available online.
Additionally, Zhaoxin and late VIA CPUs (CNQ and newer) prefer the Intel
`cpuid leaf 4' function to get cache parameters. Check whether cpuid_level
is equal to or greater than 4 to identify these CPUs and override cache values
(including L1) by calling cpu_dcp_cacheinfo(ci, 4). L3 cache information is
available only through this method.
[7 lines not shown]
Pull up following revision(s) (requested by skrll in ticket #442):
sys/arch/riscv/include/asm.h: revision 1.13
Ensure 4-byte (32bit alignment) in ENTRY_NP so that non-compressable
instructions get the correct alignment.
Pull up following revision(s) (requested by tls in ticket #441):
external/gpl3/gdb/dist/gdb/Makefile.in: revision 1.11
Fix MKCROSSGDB build from release source tars that have had permissions
whanged around
Pull up following revision(s) (requested by tls in ticket #440):
sys/arch/hppa/hppa/process_machdep.c: revision 1.19
sys/arch/hppa/include/cpu.h: revision 1.14
sys/arch/hppa/hppa/machdep.c: revision 1.28
sys/arch/hppa/hppa/machdep.h: revision 1.20
sys/arch/hppa/conf/GENERIC: revision 1.50
sys/arch/hppa/conf/files.hppa: revision 1.30
sys/arch/hppa/hppa/locore.S: revision 1.8
sys/arch/hppa/hppa/hppa_machdep.c: revision 1.34
sys/arch/hppa/hppa/fpu.c: revision 1.28
Add eager FPU switching for hppa - works around QEMU bug, and safer -
may eventually become the default.
Pull up following revision(s) (requested by tls in ticket #439):
sys/dev/pci/vioscsi.c: revision 1.37
sys/dev/pci/files.virtio: revision 1.5
Add VIOSCSI_MAX_TARGET and VIOSCSI_MAX_LUN options to reduce excessive probe time on QEMU
Pull up following revision(s) (requested by tls in ticket #438):
sys/dev/goldfish/gftty.c: revision 1.6
Make console input on virt68k 8-bit clean.
On signed-char ports you can't feed a "char *" to the tty layer. A cursory
glance for more instances of this didn't find any but it's possible.
Pull up following revision(s) (requested by tls in ticket #437):
sys/sys/systm.h: revision 1.307
sys/sys/syscall.h: revision 1.330
sys/kern/syscalls.c: revision 1.335
sys/kern/syscalls_autoload.c: revision 1.51
sys/kern/systrace_args.c: revision 1.57
sys/arch/alpha/alpha/syscall.c: revision 1.46
sys/arch/riscv/riscv/syscall.c: revision 1.7
sys/arch/mips/mips/syscall.c: revision 1.52
sys/kern/init_sysent.c: revision 1.347
sys/kern/makesyscalls.sh: revision 1.189
sys/sys/syscallargs.h: revision 1.313
Fix a longstanding bug on 64 bit ports with ABIs requiring sign-extension when returning 32-bit values in registers. Alpha, MIPS N64, RISCV.
With these ABIs it is necessary for the syscall return code to manually
sign-extend all 32-bit return values. If we don't, everything may look
just fine, until a syscall return value is spilled to the stack and then
[13 lines not shown]
Pull up following revision(s) (requested by tls in ticket #436):
sys/arch/evbmips/malta/pci/pcib.c: revision 1.29
sys/arch/evbmips/malta/dev/mainbus.c: revision 1.21
sys/arch/evbmips/malta/dev/gt.c: revision 1.20
sys/arch/evbmips/malta/malta_intr.c: revision 1.30
sys/arch/evbmips/malta/malta_bus_mem.c: revision 1.10
sys/arch/evbmips/malta/machdep.c: revision 1.51
sys/arch/evbmips/malta/malta_bus_io.c: revision 1.9
sys/arch/evbmips/conf/MALTA: revision 1.107
Address problems with MIPS Malta platform code found running under QEMU.
QEMU's "malta" system emulates a MIPS Malta with the Gallileo host bridge
and 32-bit or 64-bit CPUs of either endianness. It is one of the only
working QEMU system-level emulations that could run NetBSD with all
combinations of endianness and address size. After fixes to QEMU over the
past several years, NetBSD has been unable to use the emulated PCI bus in
big-endian and 64-bit configurations.
[36 lines not shown]
Pull up following revision(s) (requested by tls in ticket #433):
distrib/sets/sets.subr: revision 1.214
Fix MKBINUTILS=no builds. This message brought to you by the number
33,554,432 and the word 'embedded'.
