mkimage/conf/*.conf: Nix postfix minwrites customization.
This customization has been a no-op for a long time.
These config files previously had copied & pasted logic to replace
the `fifo' listener type (named pipes) by `unix' (local sockets)
/etc/postfix/master.cf, because notifying fifo listeners bumps mtime,
while notifying unix listeners does not, so changing from `fifo' to
`unix' would reduce disk writes.
But Postfix has had used `unix' for everything by default since 2.10,
released in 2012 and first shipped in NetBSD 7.0 in 2015. So this
customization has been a no-op for over a decade. Let's save readers
some pondering and just nix it.
PR/60002: Enji Cooper:
mq_open in FreeBSD and NetBSD both state that the path provided to the
target mqueue should (in the POSIX parlance of "should") start with a /.
In FreeBSD this proved to be more of an issue than NetBSD (especially
when `kyua debug` was involved). Instead of relying on a temp directory
which might exist outside the kyua sandbox, make the mqueue path
absolute to the prefix `/t_mqueue` (for consistency with the test), and
add the test PID to the path to avoid potential collisions between
back-to-back / concurrent test runs.
While here, explicitly cast the checked return value to `mqd_t`: this is
needed because some platforms--including FreeBSD--might define `mqd_t` as
an unsigned type instead of a signed type. Casting it to `mqd_t` helps
quell a valid compiler warning in that case.
catch up with hyperfb:
- track writes to SBA/DBA/BAboth
- use BINC writes to draw anti-aliased characters instead of direct fb access
- some magic number reduction
Don't use --target, because this sets both the input and the output format,
and now 2.46 does not auto-recognize the input anymore. use --output-target \
instead.
use proper string conversion for RAID level using create method
using atoi returns 0 for any non-integer passed to it. this could
configure RAID 0 in the kernel which was not intended for when one
mistakenly passes a non-integer string. uses an internal conversion
function that produces proper error message.
pullup 10 and 11
bsd.prog.mk: Respect LDSTATIC.${PROG}, not LDSTATIC.${PROG}.link.
Take two: make sure to do this only if we're actually making PIEs.
PR toolchain/60000: split debug data rules broke ${.TARGET}-dependent
flags
NFC: Cleanups only (mostly to comments).
Fix grammar, spelling, layout, and occasionally, content
of comments added in the previous commit, and even add one
or two new ones.
While here, some 80 column police action.
The only difference in the generated object files is the RCSID.
(On amd64 anyway).
PR bin/59996 - handle hidden swap list reordering
A different attempt to achieve what 2 revs back was attempting.
The swap lists must be locked (uvm_swap_data_lock) when we are traversing
the lists of swap devices, as otherwise the lists can reorder themselves
behind our back. But we cannot hold that lock when actually doing the
processing, as our process might need to page/swap to copy out data,
and doing that will also attempt to take the lock - panic (or doom).
Instead, traverse the lists with the lock held, so they are stable, but do
nothing but keep a record of all of the swapdevs (independent of their lists)
and then use this new list of swapdevs to actually do the work. The number
or identity of the swap devices cannot change during all of this, as we also
hold swap_syscall_lock which prevents any other swapctl() operations (like
adding or deleting devices) from occurring.
Once we have done that, the number of swap devices found is the number
returned from swapctl(SWAP_STATS) (provided it is no bigger than requested).
[5 lines not shown]
