When doing busdma sync ops for BUSDMA_COHERENT memory, there is no need
for cache maintenance operations, but ensure that all prior writes have
reached memory when doing a PREWRITE sync.
Do not do a cache invalidate on a PREREAD sync that is also a PREWRITE sync.
Do the cache invalidate sequence from the outermost to innermost, required
for correct operation.
Correct the sequence of busdma sync ops involved with PRE/POSTREAD syncs.
When doing a PREREAD sync of an mbuf-type dma buffer, do a writeback of
the first cacheline if the buffer start address is not on a cacheline
boundary.
Treat D keywords as identifiers in certain postfix expressions. This allows
one to, for example, access the "provider" field of a struct g_consumer,
even though "provider" is a D keyword.
PR: 169657
The module load address always needs to be included when setting the dm_*_va
fields of dt_module_t. Previously, this was only done on architectures where
kernel modules have type ET_REL; this change fixes that. As a result, symbol
name resolution in the stack() action now works properly for kernel modules
on i386.
Make isp_find_pdb_by_*() search for targets in portdb in reverse order.
Records with target_mode == 1 are allocated from the end of portdb, so it
seems logical to start search from the end not traverse whole array.
Add configuration options to override physical and UNMAP blocks geometry.
While in most cases CTL should correctly fetch those values from backing
storages, there are some initiators (like MS SQL), that may not like large
physical block sizes, even if they are true. For such cases allow override
fetched values with supported ones (like 4K).
Overhaul if_gif(4):
o convert to if_transmit;
o use rmlock to protect access to gif_softc;
o use sx lock to protect from concurrent ioctls;
o remove a lot of unneeded and duplicated code;
o remove cached route support (it won't work with concurrent io);
o style fixes.
MFC r273090:
Move memset under ifdef INET6.
MFC r273091:
Add more ifdefs. SIOC*_IN6 are defined only with INET6.
MFC r273121:
Add inet/inet6 to the dependency list. Without them if_gif is useless.
MFC r273209 by bz:
After r273087,r273090,r273091,r273121 changes to gif(4) try to fix
NOIP builds for real.
MFC r273587:
Remove redundant check and m_pullup() call.
ifdef ext2_print_inode which is not really used.
ext2_print_inode was nice to have for initial development work but
is not really used anymore. #ifdef it under a new EXT2FS_DEBUG knob
so that we don't spend time compiling it.
sed: Bounds check the file path used in the 'w' command.
Modified version of a diff from Sebastien Marie to prevent a crash found
with the afl fuzzer.
Obtained from: OpenBSD (CVS Rev. 1.37)
ZFS large block support. The default recordsize remains at 128KB.
A new tunable/sysctl variable, vfs.zfs.max_recordsize is added to
allow adjusting the permitted maximum record size, or
zfs_max_recordsize, with a default of 1MB. ZFS will not allow
setting recordsize greater than zfs_max_recordsize as a safety
belt, because larger recordsize means greater read and write
latency and more memory usage.
Please note that booting from datasets that have recordsize greater
than 128KB is not supported (but it's Okay to enable the feature on
the pool).
Limited safety belt is provided for mounted root filesystem but use
caution when using a larger value.
Illumos issue:
5027 zfs large block support
- Remove spaces from boot messages when we print the CPU ID/Family/Stepping
- Move prototypes for various functions into out of C files and into
<machine/md_var.h>.
- Reduce diffs between i386 and amd64 initcpu.c and identcpu.c files.
- Move blacklists of broken TSCs out of the printcpuinfo() function
and into the TSC probe routine.
- Merge the amd64 and i386 identcpu.c into a single x86 implementation.
Add makewhatis to ITOOLS if MK_MAN != no
This will fix installation with differing host targets in installworld, so
one can build i386/i386 on an amd64 host, then install to an i386/i386 target
Reported by: alfred
Phabric: D1280
r273803:
Filter out TESTS_SUBDIRS already added to SUBDIR instead of blindly
appending the TESTS_SUBDIRS variable to SUBDIR
Duplicate directory entries can cause unexpected side effects, like
installing the same files multiple times. This can be easily
reproduced via the following testcase prior to this commit:
SUBDIR= dir
TESTS_SUBDIRS+= dir
.include <bsd.test.mk>
Sponsored by: EMC / Isilon Storage Division
r273810:
Fix the logic inversion in the previous commit by ensuring that the matched
expression (:M) is empty, not the not matched (:N) is empty. The former case
means we have not found the TEST_SUBDIR value in SUBDIR
Reported by: rodrigc
Pointyhat to: me (did not use a clean install root)
Sponsored by: EMC / Isilon Storage Division
r273186:
Don't define rules based on PROGS if PROGS is empty.
Reviewed by: sjg, ngie
Sponsored by: EMC / Isilon Storage Division
r273873:
Use the right depend file for each program.
bsd.progs.mk generates a separate depend file for every program being
built, but then it does not properly tell each submake to use those
individual files. Properly propagate the depend file to use.
Discovered while preparing the update of atf to 0.21 and noticing that
the test programs were not being relinked to the new library.
This change is "make tinderbox" clean.
Fix smbfs to not zero out statfs f_flags field. Previously, this
made getmntinfo() return empty flags for smbfs filesystems when
called with MNT_WAIT. It's not visible with mount(8), since it uses
MNT_NOWAIT, but broke autounmount(8) operation.
PR: 195161
Sponsored by: The FreeBSD Foundation
