Switch ports using MI wsfb drivers to HAVE_XORG_SERVER_VER=120.

Xorg wsfb servers from 1.20 for dreamcast (16bpp), hp300 (8bpp), and
luna68k (1bpp) work fine even on NetBSD 9.0, and zaurus is also
confirmed working.

Discussed with mrg@ and macallan@ on tech-x11:
https://mail-index.netbsd.org/tech-x11/2020/07/thread1.html#002098

(tsutsui)

Serialize CPU hatch annoucement printfs and wait for CPUs to start before
returning from cpu_boot_secondary_processors.

(jmcneill)

Simplify IPI handling even more for now and run everything at IPL_HIGH.

(jmcneill)

Remove extra "-I" for ${X11INCS.DIX} macro.

(tsutsui)

Update Spleen wscons fonts to version 1.8.1, bringing the following
improvements:

- Center tilde vertically for all sizes, as modern fonts do

(fcambus)

Fix typos in ${X11INCS.DIX} include paths. PR/55500

(tsutsui)

Update Spleen kernel fonts to version 1.8.1, bringing the following
improvements:

- Center tilde vertically for all sizes, as modern fonts do

(fcambus)

Fix confusion between ipi bitmask and mbox register bit assignments.

(jmcneill)

Move exc_step definition under MIPS64_OCTEON and use it in the zeroising
memset as length for correctness.

(skrll)

clear all interrupts, not just those we expect from the hostintmask.

this removes the final hard hang i have seen in pinebookpro wifi,
though one may still need to 'ifconfig bwfm0 down up' occasionally,
so we still have bugs to fix here (the hang is usually associated
with 'checksum error' from bwfm/sdio.)

(mrg)

only ask for SDPCMD_INTSTATUS_HMB_SW_MASK and SDPCMD_INTSTATUS_CHIPACTIVE
interrupts, not all of them.  we only ack these ones.

mostly fixes pinebookpro wifi hard hangs.  still is problematic and can
trigger interrupt storm that appears as a hard hang without NET_MPSAFE,
and a follow up, less clearly right, change will reduce that to a soft
hang of the interface that can be cleared with 'ifconfig bwfm0 down up',
and even often recovers itself now.

(mrg)

Revert previous, to unbreak the build (NVMM declares the macro too).

There are hundreds of MSRs, we're not going to list them all, especially
when the majority are unused.

(maxv)

Less magic numbers.

(simonb)

Remove check/limit for only 2 cores.
Use MIPS_EBASE_CPUNUM instead of magic numbers.

(simonb)

Use MIPS_EBASE_CPUNUM instead of magic numbers.

(simonb)

Expose MIPS_EBASE_CPUNUM.

(simonb)

Expose the width of the MIPS_EBASE_CPUNUM bitfield for asm code.

(simonb)

Use COPY_SYMTAB to get kernel symbols.

(uwe)

Add an extra bitfield in MIPS_COP_0_EBASE.

(simonb)

Implement cngetc to be able to talk to DDB.

(uwe)

Add placeholders/boilerplate for NextVoD.

STMicroelectronics support needs to be written for both the CPU and
devices.  This config and related files is absolute bare bones to get
a scaffolding kernel to compile to actually work on that support.
Don't hold your breath.

(uwe)

make(1): remove unnecessary UNCONST

(rillig)

make(1): make ampersand in ${VAR:from=to&} an ordinary character

In SysV substitutions, wildcards are expressed with % instead of &.  The
& is not mentioned in the manual page, and having another wildcard for
the whole word would be such an obscure feature that not even pkgsrc uses
it.  The easiest way to discover this feature had been to read the source
code of make(1) or to use a fuzzer and accidentally stumble upon this
edge case.

(rillig)

make(1): don't call VarSubstitute unnecessarily

(rillig)

make(1): remove wrong VarPattern from ApplyModifier_Exclam

(rillig)

make(1): remove wrong VarPattern from ApplyModifier_Assign

(rillig)

make(1): add test for assignment operators in conditions

(rillig)

make(1): flatten ApplyModifier_Assign

(rillig)

make(1): add tests for the ::= modifiers

(rillig)

make(1): fix date in comment in unit test

(rillig)

make(1): add tests for maximum number of capturing groups in :C

(rillig)

make(1): fix off-by-one error in :C modifier

Previously this off-by-one error had invoked undefined behavior.
Until today there was no corresponding unit test though.

(rillig)

Sync with xorg-server.old: Remove #ifdef XKB conditional.

> It has been mandatory on Xorg:
>  https://cgit.freedesktop.org/xorg/xserver/commit/?id=40877c6680863bd6a8475f2bb0c54df55bcf0b0e

(tsutsui)

Remove #ifdef XKB conditional.

It has been mandatory on Xorg:
https://cgit.freedesktop.org/xorg/xserver/commit/?id=40877c6680863bd6a8475f2bb0c54df55bcf0b0e

(tsutsui)

make(1): fix double free introduced this morning in r1.258

(rillig)

make(1): remove redundant loop control variable

(rillig)

make(1): rename ApplyModifiersState.flags to eflags

There are just too many variables called "flags" in this file, and their
constants must not accidentally overlap.

(rillig)

make(1): demonstrate off-by-one bug in :C modifier

(rillig)

make(1): clean up VarLoop structure

(rillig)

make(1): remove unused parameter from ParseModifierPart

(rillig)

add #ifdef __HAVE_PCI_MSI_MSIX so this still compiles with NO_PCI_MSI_MSIX

(jdolecek)

add definition for MSR_IA32_FEATURE_CONTROL, just for information

(jdolecek)

make(1): rename VarGetPattern to ParseModifierPart

(rillig)

make(1): eliminate unused local variables

(rillig)

make(1): eliminate macro from VarGetPattern

(rillig)

Nix trailing whitespace.

(riastradh)

Expand on importance of not using fpu for crypto if there's no fpu.

(riastradh)

make(1): add test for stray dollar signs in :@ modifier

(rillig)

Remove unused Pp. Sort SEE ALSO.

(wiz)

sync with reality

(maxv)

Compile USER_LDT by default, but, put it behind a privileged sysctl that
defaults to disabled. To enable:

# sysctl -w machdep.user_ldt=1

(maxv)

for Xen MSI, fallback to INTx when PHYSDEVOP_map_pirq fails for the device

apparently Xen requires VT-d to be enabled in BIOS for PHYSDEVOP_map_pirq
to work, this change makes it work on systems with VT-d disabled or missing

adresses the panic part of PR port-xen/55285 by Patrick Welche

(jdolecek)

make(1): add test for ampersand in SysV substitution modifier

(rillig)

make(1): replace ApplyModifiersState field with local variable

(rillig)

we're already in an #ifdef USER_LDT block, so no need to #ifdef again

(maxv)

don't include opt_user_ldt.h when it is not needed

(maxv)

make(1): rename flags for :S and :C pattern matching

Renaming these flags makes it obvious that VarPattern is used in a few
places where it doesn't belong.  These will be fixed in follow-up
commits.

(rillig)

make(1): convert VarFind flags to typed enum

(rillig)

make(1): clean up unnecessary snprintf and multi-line function calls

(rillig)

make(1): rename Varf_Flags to VarEvalFlags

In var.c there are lots of different flag types.  To make any accidental
mixture obvious, each flag group gets its own prefix.

The only flag group that is visible outside of var.c is concerned with
evaluating variables, therefore the "e", which replaces the former "f"
that probably just meant "flag".

(rillig)

Fix RPI boot which needs more early stack

XXX centralise INIT_ARM_STACK_{SHIFT,SIZE}
XXX how to protect against this breaking again?

(skrll)

make(1): clean up handling of :? modifier

The VarPattern type had nothing to do with parsing conditions.

(rillig)

make(1): clean up code in var.c

* use canonical comparison to zero in more cases
* omit unnecessary snprintf for string literal concatenation
* omit a few redundant parentheses and braces

(rillig)

A little KNF.

(simonb)

make(1): move SysV string matching to var.c

This kind of string matching is only used in variable modifiers, and only
if this feature is enabled by SYSVVARSUB.

(rillig)

make(1): add more tests for :ts modifier

(rillig)

KNF whitespace nits.

(simonb)

Reserve space for all CPU cores (incorrect logic).
Reserve a full page for each core, not incorrect 4096 (from skrll@).

(simonb)

Revert most of ad's movs/stos change. Instead do a lot simpler: declare
svs_quad_copy() used by SVS only, with no need for instrumentation, because
SVS is disabled when sanitizers are on.

(maxv)

fix build with clang/llvm.

clang aarch64 assembler doesn't accept optional number of lanes of vector register.
(but ARMARM says that an assembler must accept it)

(ryo)

fix build error with LLVM.

(ryo)

make(1): demonstrate inconsistent :ts behavior among modifiers

(rillig)

Switch to fpu_kern_enter/leave, to prevent clobbering, now that the kernel
itself uses the fpu.

(maxv)

The TLB flush IPIs do not respect the IPL, so enforcing IPL_HIGH has no
effect. Disable interrupts earlier instead. This prevents a possible race
against such IPIs.

(maxv)

Call fdtbus_intr_init() during startup.

(simonb)

Add -dh for DEBUG_HASH

Allow tracking of max chain length, to see how well the hash
tables are working.
Pull the actual hash operation into a marco so it can be
easily changed - for experimenting.

The current hash, is pretty good.

Reviewed by: christos

(sjg)

Now that the IDT is per-CPU, it must be saved/restored on each CPU
independently.

(maxv)

missing e.g.

(nia)

more hardware

(nia)

clarify

(nia)

Initial evbarm upgrade instructions - need work

(nia)

Pull 1bpp server fixes from xorg-server.old:

https://mail-index.netbsd.org/source-changes/2020/07/18/msg119488.html
> Fix 1bpp Xservers on "whitePixel=0, blackPixel=1" VRAMs.
>
> - Don't override pScreen->blackPixel and pScreen->whitePixel
>  (set in MD server Init functions per -filpPixels option)
>  on 1bpp servers in merged fbSetupScreen() (merged one
>  from cfbSetupScrenn and mfbSetupScreen() in old xsrc/xfree)
> - Pull mfbCreateColormap() function from old xsrc/xfree tree
>  and use it on 1bpp servers

Confirmed on luna68k 1bpp Xorg 1.20 server using xf86-video-wsfb
with "-flipPixels" option.

(tsutsui)

more contents

(nia)

Fix 1bpp Xservers on "whitePixel=0, blackPixel=1" VRAMs.

- Don't override pScreen->blackPixel and pScreen->whitePixel
  (set in MD server Init functions per -filpPixels option)
  on 1bpp servers in merged fbSetupScreen() (merged one
  from cfbSetupScrenn and mfbSetupScreen() in old xsrc/xfree)
- Pull mfbCreateColormap() function from old xsrc/xfree tree
  and use it on 1bpp servers

Required by forthcoming X11R7'fied ancient Xsun and XsunMono servers:
https://mail-index.netbsd.org/port-sun3/2020/07/17/msg000157.html

Without these fixes, mouse pointers and twm window titlebar etc. are
rendered inverted.

(tsutsui)

safe estimation of required storage

(nia)

more modernization of evbarm install instructions

(nia)

Always call pmap_segtab_activate in pmap_activate.  pmap_segtab_activate
does the right thing if called with non-curlwp.

(skrll)

Xorg 1.20 servers require libXfont2 instead of libXfont.

Per the following upstream commit:
https://cgit.freedesktop.org/xorg/xserver/commit/?id=05a793f5b3c40747d5a92a076def7f4fb673c7e7
> dix: Switch to the libXfont2 API (v2)

Now Xorg 1.20 based X68k server builds and confirmed working on XM6i.

(tsutsui)

Add a stub for miinitext of GLX.

As per upstream commits for Xnest:
https://cgit.freedesktop.org/xorg/xserver/commit/?id=67c303fff303f94b62f03a76de97116c6ebcfda9

Note GLXEXT is defined as 1 in xorg-server/include/xorg-server.h
(i.e. /usr/X11R7/include/xorg/xorg-server.h) so no proper way to
disable it on building MD servers per ${MACHINE} basis.

(tsutsui)

Update for removal of AddEnabledDevice and RemoveEnabledDevice.

Use the NotifyFd API instead as per upstream commits:
https://cgit.freedesktop.org/xorg/xserver/commit/?id=55c2e1a3aa587c58a74066724e11e30b3df267b8
https://cgit.freedesktop.org/xorg/xserver/commit/?id=be5a513fee6cbf29ef7570e57eb0436d70fbd88c
https://cgit.freedesktop.org/xorg/xserver/commit/?id=60a91031d13e4d29c383087120e318f6b528b6e5
https://cgit.freedesktop.org/xorg/xserver/commit/?id=7def2fea30060d47780dc1eedc91fada5ae1934f

(tsutsui)

Sync with xorg-server.old: No need to link miext/libshadow.a.

(tsutsui)

No need to link miext/libshadow.a.

(tsutsui)

Add missing linebreak.

Reported by Anon.

(wiz)

Add BIOCSETWF32 (bpf)

(jmcneill)

Pull changes of x68kInit.c rev 1.3 from xorg-server.old.

https://mail-index.netbsd.org/source-changes/2020/07/18/msg119473.html
> Use AllocDevicePair() to initialize input devices in InitInput().

Untested, but I'll re-visit on migrating to HAVE_XORG_SERVER_VER=120.

(tsutsui)

Use AllocDevicePair() to initialize input devices in InitInput().

This should have been done on the initial X11R7'fied imoprt as
upstream xnest did:
https://cgit.freedesktop.org/xorg/xserver/commit?id=32ce5c83a504d2cafd2b19ed628b40032ef347b3

Tested and no visible regression on XM6i.

(tsutsui)

ftp: add -? for help. improve synopsis

Add -? to display usage synopsis and help to stdout.
This allows for "ftp -? | less", which is more user friendly.
Errors still show usage to stderr.
Consistency improvements in some usage text.

(lukem)

Remove 2 CPU limit in OCTEON interrupt controller driver.

(jmcneill)

remove explicit irqack() call for unexpected interrupt; this happens
normally during boot when running identify for the drive as polled
command, and fixes hard hang caused by the irqack() on a sun machine

the unexpected interrupt happens because cmdide doesn't respect
WDCTL_IDS bit and triggers interrupt also for polled commands

also reclassify the "bogus intr" as aprint_verbose() to avoid noise
on boot

discussed and tested by Martin Husemann

(jdolecek)

Simplify IPI handling and change IPLs of IPI_HALT, IPI_XCALL, and
IPI_GENERIC from IPL_SCHED to IPL_HIGH.

(jmcneill)

Remove trailing whitespace.

(wiz)

Cleanup handling of multiple banks.

(jmcneill)

libc: Added ppoll() wrapper around native pollts(2)

(kamil)

Enable support for ppoll()

(kamil)

Add ppoll() a compatibility wrapper around pollts(2)

Submitted by Apurva Nandan.

(kamil)

man5/changelist.5 does not depend on MKPOSTFIX

(kamil)

Create an "octeon.img.gz" bootable image for evbmips64-eb release builds.

(jmcneill)

ERLITE -> OCTEON

(jmcneill)

ERLITE -> OCTEON

(jmcneill)

Add wedge support

(jmcneill)

don't install xen/xenio.h and xen/xenio3.h anymore, xentools don't use it

PR port-xen/52874

(jdolecek)

if bnx_tx_encap() fails because mbuf is too fragmented or too long,
drop the mbuf instead of wedging the TX queue forever; found by code inspection

this is quite unlikely scenario since it requires mbuf chain consisting of
more than 8 frags and m_defrag() failing, so probably unrelated to PR kern/47229

(jdolecek)

make sure bnx_alloc_pkts() always clears BNX_ALLOC_PKTS_FLAG and IFF_OACTIVE,
even on dmamap allocation failure; remove unnecessary IFF_UP condition

(jdolecek)

remove mistakely committed in-progress code from previous commit

(jdolecek)

re-enable MSI/MSI-X, the TX timeouts were caused by the IFF_OACTIVE handling,
which was fixed in previous revision

(jdolecek)

Don't use a reserved value for the USB endian CSR selects and
enable octxhci_uctl_init().  Between these, USB works without
needing a "usb start" from u-boot.

XXX: Note the port power enable goop is still disabled until
we get a GPIO driver.

(simonb)

KNF. 80 cols, use tab.

(ryo)

Add options PMAPBOOT_DEBUG to dump TTBR when pmapboot_enter().
Formerly DEBUG_MMU in locore.S, but there was a bit of confusion.

(ryo)

Note x86 per-cpu IDTs change

(yamaguchi)

Tweak a bit and add some comments.

Same object code is generated.

(uwe)

Do not recycle loop variable for unrealated purposes.

It just obfuscates the code.  Same object code is generated.

(uwe)

Add USB3 support.

(jmcneill)

Support 128 IRQs instead of 64. This is icky and needs to be cleaned up.

(jmcneill)

Add of_find_bycompat helper to search a tree for a node by compat string.

(jmcneill)

Add driver for Cavium Interrupt Bus.

(jmcneill)

Initialize FDT console device

(jmcneill)

Use uart0 for stdout-path

(jmcneill)

Add kernel option FDT_DEFAULT_STDOUT_PATH to set default (fallback)
stdout-path

(jmcneill)

Make consinit actually work

(jmcneill)

Don't accept negative value.

Reported-by: syzbot+e71a77402d6668f1868d@syzkaller.appspotmail.com

(msaitoh)

need to clear IFF_OACTIVE after allocating more pkts in order to actually
start processing the queue in bnx_start()

(jdolecek)

disable MSI/MSI-X for time being, it causes interrupt timeouts and needs
to be investigated before enabling this again

(jdolecek)

make sure the alignment of ciss_cmd matches KASSERT() in ciss_cmd()

(jdolecek)

revert the conversion to STAILQ, it broke execution of commands

(jdolecek)

Fix variable name for BUS_DMA_DEBUG.

(simonb)

FDT support for Cavium OCTEON MIPS SoCs. WIP.

(jmcneill)

Add glue for 8250 compatible UARTs.

(jmcneill)

Allow kernels to link with no console drivers

(jmcneill)

pmapboot_enter simplication
- bootpage_alloc in asm becomes pmapboot_pagealloc in C
- PMAPBOOT_ENTER_NOBLOCK is removed as it's not used
- PMAPBOOT_ENTER_NOOVERWRITE is removed as it's now always on
- physpage_allocator argument is removed as it's always
  pmapboot_pagealloc
- Support for EARLYCONS without CONSADDR is removed so that the identity
  map for CONSADDR is always known.

For the assembly files:
2 files changed, 40 insertions(+), 89 deletions(-)

LGTM ryo

(skrll)

+craptacular

(originally suggested by pgoyette)

(dholland)

Set PCI_COMMAND_MASTER_ENABLE and PCI_COMMAND_MEM_ENABLE
to activate the pci devices

This configuration is needed when BIOS or UEFI do not make them set.

(yamaguchi)

Try to improve markup for better PostScript output.

(uwe)

Document unbound and unbound_chrootdir.

(leot)

MKBSDTAR is yes by default

(leot)

Do not use "[...]", just "..." is enough.

Conventionally the ellipsis already expresses optional repetition,
e.g. .Ar without arguments produces "file ...".

(uwe)

's/blacklistd/blocklistd/'. Note also blocklistd_flags.

(jruoho)

Note modules.

(jruoho)

Remove now-extraneous Op since we already have Oo and Oc to enclose
the Ar port.

(pgoyette)

Identify SDHC 4.1 and 4.2. From {DragonFly,Free}BSD.

(msaitoh)

Fix typo. Use PRIxPADDR rather than casting.

(rin)

g/c unused sc_channel_raw, sc_adapter_raw

(jdolecek)

Document optional speed argument to consdev

(kim)

Let consdev command also set speed

Adapted from PR install/55490 by Sunil Nimmagadda

(kim)

Rename the evbmips ERLITE kernel to OCTEON now that it supports more
than just the EdgeRouter Lite.

(simonb)

Add NetBSD RCSID. No functional changes.

(rin)

Now, FPU emulation for booke and ibm4xx works fine at a level where
all the related ATF tests pass correctly. However, there still remain
problems:

- FEX and VX bits for FPSCR cannot be modified by mcrfs, mtfsf{,i},
  and mtfsb[01].
- Invalid operations should be treated differently depending on
  FPSCR[VE].

Therefore, comment them in order not to be forgotten.

No binary changes.

(rin)

Try to fix FPSCR bits in the end of emulation:

- FPSCR[FEX] is not a sticky bit.
- Turn on FPSCR[FEX] if the emulated instruction causes invalid operation,
  and invalid operation exception is not masked out.
- FPSCR[VX] is not a sticky bit, however it should be set when at least
  one of FPSCR[VXfoo] bits (they are sticky!) is set.
- FPSCR[FX] is a sticky bit, and it should be set if FPSCR is modified by
  instructions other than mtfsf{,i}.

(rin)

Set ksi_code correctly via fpu_get_fault_code() for SIGFPE.

(rin)

Expose fpu_get_fault_code() even if !PPC_HAVE_FPU, and
adjust it to systems without FPU.

(rin)

Do not raise SIGFPE unless MSR[FE0] or MSR[FE1] is set via fenv(3).

(rin)

For booke and ibm4xx, emulate m[ft]msr in user mode, in the same
manner as oea.

Now, user process can decide by itself whether floating-point
exception triggers SIGFPE or not via fenv(3).

(rin)

Factor out emulation code for m[ft]msr in user mode from oea, and
adjust it for systems without FPU.

Now, it can be used from booke and ibm4xx in order to support fenv(3).

(rin)

ftp.1: don't wrap "[[user@]host [port]]"

(lukem)

Treat trap instruction from userland correctly in EXC_PGM handler;
raise SIGTRAP with TRAP_BRKPT instead of SIGILL.

(rin)

FPSCR[FEX] is not a sticky bit; it is always cleared when read from
userland via mffs on real hardware.

(rin)

PR port-powerpc/55425

Update comment; FPU emulation seems to work just fine now. However,
FPU-optimized code should still be avoided for better performance,
if FPU is not present.

(rin)

PR port-powerpc/55425

Fix emulation for mtfsf; source register is frB here.

Now, userland processes successfully change rounding mode, by which
FPU-optimized code in OpenSSL works just fine as far as I can see.

(rin)

Rename emulated_opcode() to emulate_privileged() for clarity.
No functional changes.

(rin)

For trap instruction, ksi_code should be TRAP_BRKPT not TRAP_TRACE.

(rin)

Do not set ksi->ksi_addr twice. No functional changes.

(rin)

Do not raise divide-by-zero exception when dividend is zero.

(rin)

Do not use curlwp twice. No functional changes.

(rin)

Remove old workaround foe cache problem on ibm4xx.
The problem seems gone already.

(rin)

Add missing note about Intel I219 LM10-LM15 and V10-V14.

(msaitoh)

G.C. comment.

(msaitoh)

Fixup OpenSimpleReadFile usage:
- The "EFI_DEVICE_PATH **FilePath" parameter can change, so do not free
  the output. This was causing crashes on U-Boot when attempting to load
  boot.cfg, even if it didn't exist.
- Allocate the SIMPLE_READ_FILE in advance and store a pointer to it in
  struct open_file.

(jmcneill)

note ciss(4) adapter match update

(jdolecek)

update list of matched adapters

(jdolecek)

match newer HP Smart Array controllers, list from FreeBSD

(jdolecek)

fix __arraycount() parameter

(jdolecek)

make ciss_pci_devices[] more compact

(jdolecek)

regen

(jdolecek)

add newer ciss(4) devices, list obtained from FreeBSD

(jdolecek)

mark diagused variable as such

fixing non DIAGNOSTIC builds

(para)

make bnx_wk (used to trigger bnx_alloc_pkts()) part of softc instead
of using a static variable, so it's independant for each adapter

(jdolecek)

remove call to bnx_alloc_pkts() from bnx_init_tx_chain() - it can
be called from softint context via bnx_watchdog(), and bnx_alloc_pkts()
dmamap allocation can't run in (soft)interrupt context

(jdolecek)

remove CISS_NO_INTERRUPT_HACK, discussed with joerg@ and spz@

(jdolecek)