This prevents the following race condition.
1. Thread-A: calls virtio_dequeue_commit() and
puts a slot into free descriptor chain in vq_free_slot()
2. Thread-B: calls virtio_enqueue_prep() and get the slot stored by Thread-A
3. Thread-B: calls virtio_enqueue_reserve() and
changes desc_base and desc_free_idx for the slot
4. Thread-A: changes the same members updated by Thread-B
reported by hannken, thanks.
basename(3) is invoked more than once on different paths. Retain the
correct name for the utility itself on error. Addresses PR bin/57355
from RVP, who provided the patch.
EFI bootstrap assumes it can copy the amd64 kernel to its ELF load
address (that is KERNTEXTOFF - KERNBASE = 0x200000), but it can
clash with previous UEFI memory allocation, as described here:
http://mail-index.netbsd.org/tech-kern/2023/04/07/msg028833.html
This change adds a reloc keyword for controling where the EFI
boostrap will copy the kernel image. Possible values are:
default - the default and prior behavior, copy at 0x200000.
none - do not copy and use the kernel image where it was loaded.
address - specify an explicit address where to copy the kernel.
This comes with an amd64 kernel patch that makes it self-relocatable.
It first discover where it was loaded in memory, and if this is
different than the expected 0x200000, hhe the kernel relocates
itself and start over at the right address.
On ARM ports, the EFI boot loader handles a boot.cfg file if it
exists, but one is not installed by default. In contrast, both the
i386 and amd64 ports do install boot.cfg by default. Bring these
ports into agreement by installing boot.cfg on ARM ports with EFI boot
loaders.
represented by the Rock Ridge extensions would actually differ. We would
omit the record for an all-upper-case directory name, however Linux (and
perhaps other operating systems) map names with no NM record to
lowercase.
This affected only directories, as file names have an implicit ";1"
version number appended and thus always differ. To solve, just emit NM
records for all entries other than DOT and DOTDOT .
We could continue to omit the NM record for directories that would avoid
mapping (for example, one named 1234.567) but this does not seem worth
the complexity.
From FreeBSD https://reviews.freebsd.org/D39258
incorrect, and timestamps were written in the wrong order.
See RRIP 4.1.6 Description of the "TF" System Use Entry for details.
From: https://reviews.freebsd.org/D39221
bits of hash. Unfortunately due to an implementation bu and the fact that
the return type is unsigned long which is 64 bits in LP64, this can fail
in some cases: "\xff\x0f\x0f\x0f\x0f\x0f\x12". See:
https://maskray.me/blog/2023-04-12-elf-hash-function
From Ed Maste @ FreeBSD:
https://cgit.freebsd.org/src/commit/\
?id=29e3a06510823edbb91667d21f530d3ec778116d
Need to write Unit Tests for this.
(Very long worms in a smallish window could result in the max number of
worms being 0...)
While here (also suggested by RVP) seed the random number generator,
also add a -S option to set the seed (note: while this is documented
in the updated man page, it does not appear in the usage message in
case of an error ... not likely to be used often enough to include there).
Also some minor improvements suggested by RVP:
delete the prototype for main()
exit curses mode before abort() (which should not happen, but...)
no need to return (->exit) after abort() as modern abort() can
never return.
In addition, check for extraneous (ignored) (non-option) args.
Check for absurdly small or big screens (the worm placement algorithm
doesn't work well for lines of columns < 3, and the abort() mentioned
above actually happens if one of those is == 1).
More flavours of worms added.
Some minor man page wording improvements.
Function 'leaveblock' was generating "break" label before removing
variables from the closing block. If 'createlabel' created a 'close'
instruction (which it did when matching a goto/break that exited
the scope of an upvalue), that instruction would use the wrong level.
rillig