protected with spec_io_enter()/spec_io_exit() so there is no need
to force specific vnode locking.
Set locking requirement to '= = =' (unchanged, locked or unlocked).
PR kern/57450 (unplugging hung USB disk triggers panic via _vstate_assert)
Add more variables from build.sh, BUILDING, and various
Makefile's documentation:
CDEXTRA CONFIGOPTS KERNARCHDIR KERNCONFDIR
KERNOBJDIR KERNSRCDIR LOCALTIME
Default sentences now consistently end with ".".
Consistency tweaks in "otherwise".
Editorial tweaks.
Add more variables from BUILDING:
INSTALLBOOT_BOARDS INSTALLWORLDDIR NOCLEANDIR
NODISTRIBDIRS NOINCLUDES.
Add build.sh related notes (from BUILDING).
Clarify "target" as "target architecture" where appropriate
(versus a make target where that's obvious from context.)
Move all obsolete and deprecated variables to
new "OBSOLETE VARIABLES" section at the end.
Add some per-variable editorial improvements from BUILDING.
Change sub headings to headings; it's easier to read.
(I don't know why mdoc doesn't allow blank lines around .Ss)
Add more messages for various sets operations.
Add ${_MKMSG_CREATE} to bsd.own.mk.
Enhance maketars to be a bit more verbose even with -q,
to make it more obvious why certain operations may be slow.
May help debugging PR 47562.
Sync user variables from mk.conf(5), sorted.
(Arguably this could just refer to mk.conf(5)
or share/man/man5/mk.conf.5 and remove the copypasta.)
Document NOxxx overrides in own subsection.
Unexpand tabs, tweak after.
MKDTB and MKPIE are .YorN.
MKX11=yes wants MKINET6=yes.
MKLINKLIB deserves a dash list (like bsd.README).
Expand per bsd.README: MKSTATICPIC, MKUNPRIVED,
MKX11, MKZFS, USE_LIBCSANITIZER, USE_SSP
Change the width of the variable lists to 14n (from 15n)
so that when the lists are rendered and then left aligned,
the column is 16 characters (aka 2 tabs) which makes
copypasta to bsd.README easier.
Rename compiler-warning-disable variables from
GCC_NO_warning
to
CC_WNO_warning
where warning is the full warning name as used by the compiler.
GCC_NO_IMPLICIT_FALLTHRU is CC_WNO_IMPLICIT_FALLTHROUGH
Using the convention CC_compilerflag, where compilerflag
is based on the full compiler flag name.
Provide a single variable
CC_WNO_ADDRESS_OF_PACKED_MEMBER
with options for both clang and gcc, to replace
CLANG_NO_ADDR_OF_PACKED_MEMBER
CC_NO_ADDR_OF_PACKED_MEMBER
GCC_NO_ADDR_OF_PACKED_MEMBER
Using the convention CC_compilerflag, where compilerflag
is based on the full compiler flag name.
MKLLD, MKLLDB, MKMCLINKER aren't used in the build,
so remove them.
These were obsoleted by joerg@ on 2019-11-11
with the update of LLVM to 10.0.0git (01f3a59).
For MK* variables (in _MKVARS.yes), default to
MK*.${MACHINE} if MK*.${MACHINE_ARCH} isn't defined.
Equivalent to _MKVARS.no change by mrg@ in rev 1.1065 on 2018-06-16.
Create _NOVARS for the list of NO* variables,
for consistency with _MKVARS.yes and _MKVARS.no
Format _NOVARS, _MKVARS.yes, and _MKVARS.no with
one entry per line, and sort the lists.
Simplifies review and adding new entries.
Per suggestion from riastradh@
Line up some variable assignments more consistently.
It's expressed as an integer, and is also returned as an integer type
should one retrieve it via sysctl(3) (as this man page already
correctly documents in the table above its description).
Add "CC_" prefix to NO_ADDR_OF_PACKED_MEMBER.
We now have the following convention:
GCC_NO_*: Disable specific warnings from GCC.
CLANG_NO_*: Disable specific warnings from Clang.
CC_NO_*: Disable specific warnings from both Clang and GCC.
This reserves "NO_*" Makefile variables to disable other features.
Some space->tab fixes.
List the NO variables in order, leaving enough room to add more entries,
not splitting the next letter after "NO" across a line.
No functional change; just prettier and easier to review.
Remove the leftover remnants of deprecated USE_COMBINE and NOCOMBINE,
as gcc -combine has been EOL for many years.
Completes the work done by maya@ on 2019-05-10 in bsd.kmodule.mk rev 1.61.
Hasn't worked for at least 20 years (and never in the cross-build environment,
nor did it work with parallel make), and has never been needed in NetBSD as we
don't have the older BSD programs (pascal, pre-nvi ex) that needed mkstr/xstr
on PDP-11.
PR toolchain/35964
This allows a caller to grab some data, consume part of it, and
atomically update the uio with only the amount it consumed. This
way, the caller can use a buffer of a size that doesn't depend on how
much it will actually consume, which it may not know in advance --
e.g., because it depends on how much an underlying hardware tty
device will accept before it decides it has had too much.
Proposed on tech-kern:
https://mail-index.netbsd.org/tech-kern/2023/05/09/msg028883.html
(Opinions were divided between `uioadvance' and `uioskip'. I stuck
with `uioskip' because that was less work for me.)
man0 creates files, so write to the the OBJDIR instead of the
source directory.
Use TOOL_ appropriately.
Add MAKEVERBOSE support.
Add separate rules to generate:
- all.files: the list of all man files to process (based on mkman)
- manNpages.files: per-section files from all.files. (based on mkman)
- manNpages.ps: per-section .ps from manNpages.files, with
per-file MAKEVERBOSE messages to stderr (as before, just prettier).
- manNpages.pdf: per-section .pdf from manNpages.ps.
Remove the now-obsolete mkman script.
Add TOOL_GROPS and _MKSHMSG_FORMAT to <bsd.own.mk>
OpenSSL protects the symbols in the shared object via a linker
script, but programs and tests need those protected symbols, so
they either must link against the archive, or I need to expose the
symbols. For now, I chose to expose the symbols. Note that we
cannot use -Bstatic -Lpath -llibrary -Bdynamic because this seems
to be broken with collect2 and so we use the path directly.
Instead of telling that x86/boot(8) reloc command needs a kernel able
to self relocate, we can tell it needs a kernel built with the
SELFRELOC option. This keeps the reader from wondering what could
make a kernel able to self relocate.
It was formerly known as mntzz, which Alain released three years ago.
Since then, the ZZ9000 had several firmware updates which changed some
hardware interface details rendering the former driver non functional in some
aspects. Also the audio card plug-in ZZ9000AX became available from MNT
Research. Considering the major rewrite of the driver in some areas he
decided to rename it to zz9k(9). The driver consists of several sub-drivers
each addressing different functionality of the MNT ZZ9000 & ZZ9000AX combo
card.
zz9k* is the main card driver and need to be enabled if any of the sub-driver
is enabled. zz9k on its own is not very useful, it only provides a common
zz9kbus for the other sub-drivers to connect to, so only enable it if one of
the zz9k sub-drivers are enabled.
zzfb* represents the graphics driver for the boot console and the dumb
framebuffer for X11 based on WSCONS. ZZFB_CONSOLE option enables the ZZ9000
to become the boot console.
zz* represents the ethernet interface of the ZZ9000. It basically works but
is considered experimental.
zzax* represents the ZZ9000AX audio card driver, audio* attaches to it to
provide audio output and input functionality. The driver is not functional
yet.
zzusb* represents the ZZ9000AX usb driver. It was not implemented yet and
probably never will.
Otherwise if it is empty before we were adding an optargless -d, which
ended up using the next argument (a -Isomething) after the recent change
to xlint.c related to sysroot. This broke using the native tools make
with the cross toolchain (for example typing make in /usr/src/bin/ls)
to build and install a single program.
We were supposed to keep the option list in x86/boot(8), x86/dosoot(8)
and x86/pxeboot(8) in sync, but it did not happen, hence it may work
better with all the options in x86/boot(8). Also add the undocumented
UEFI boot options.
Sorting the variables by size would be platform-dependent and thus is
not possible.
Sorting the variables alphabetically may or may not make the code easier
to read, and the example given below that rule doesn't follow it,
otherwise the correct order would be 'eight, eleven, nine, ten,
thirteen, twelve'.
https://mail-index.netbsd.org/tech-userlevel/2023/04/11/msg013749.html
hannken