i386/loader: force 4k linker max page size
Force the BIOS loaders to use the standard 4k linker max page size
so that they fit in the limited space even if someone has set
a larger default linker max page size via /etc/src.conf.
Reviewed by: imp
Sponsored by: Netflix
bsnmpd: append to LDFLAGS rather than replacing it
Sponsored by: Netflix
trim: add missing include of src.opts.mk, remove unneeded LDFLAGS
Sponsored by: Netflix
beep: add missing include of src.opts.mk, use LIBADD instead of LDFLAGS
Sponsored by: Netflix
UFS: panic rather than forcibly unmount the root fs
If the root fs is forcibly unmounted then basically every process
will die with a SEGV as soon as it tries to run again because libc.so
is gone, which leaves the system basically hung. It seems better
to just panic instead, so let's do that.
PR: 276944
Requested-by: karels
Reviewed-by: imp, mckusick, karels
Sponsored-by: Netflix
Differential Revision: https://reviews.freebsd.org/D41387
(cherry picked from commit 60a41168d195cc09f7351c2c2bcafc389ed52406)
x86/ucode: add support for early loading of CPU ucode on AMD.
Sponsored by: Netflix
Reviewed by: markj
Differential Revision: https://reviews.freebsd.org/D43318
x86/ucode: add const where appropriate
Sponsored by: Netflix
Reviewed by: markj
Differential Revision: https://reviews.freebsd.org/D43865
mpi3mr: add missing terminator entry to mpi3mr_identifiers array
Reported by: KASAN
Sponsored by: Netflix
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D43519
prf: add setting of msgbuftrigger in paths where it is missing
Logging things to msgbuf is supposed to set msgbuftrigger so that
syslogd will notice that there is more to read from /dev/klog,
but several paths do not do that. Add the missing trigger.
Reviewed by: imp, kib
Differential Revision: https://reviews.freebsd.org/D42173
UFS: panic rather than forcibly unmount the root fs
If the root fs is forcibly unmounted then basically every process
will die with a SEGV as soon as it tries to run again because libc.so
is gone, which leaves the system basically hung. It seems better
to just panic instead, so let's do that.
Requested-by: karels
Reviewed-by: imp, mckusick, karels
Sponsored-by: Netflix
Differential Revision: https://reviews.freebsd.org/D41387
ffs: restore backward compatibility of newfs and makefs with older binaries
The previous change to CGSIZE had the unintended side-effect of allowing
newfs and makefs to create file systems that would fail validation when
examined by older commands and kernels, by allowing newfs/makefs to pack
slightly more blocks into a CG than those older binaries think is valid.
Fix this by having newfs/makefs artificially restrict the number of blocks
in a CG to the slightly smaller value that those older binaries will accept.
The validation code will continue to accept the slightly larger value
that the current newfs/makefs (before this change) could create.
Fixes: 0a6e34e950cd5889122a199c34519b67569be9cc
Reviewed by: mckusick
MFC after: 3 days
Sponsored by: Netflix
(cherry picked from commit d464a7698de8fe18530ea65fac90dce56b860a59)
ffs: restore backward compatibility of newfs and makefs with older binaries
The previous change to CGSIZE had the unintended side-effect of allowing
newfs and makefs to create file systems that would fail validation when
examined by older commands and kernels, by allowing newfs/makefs to pack
slightly more blocks into a CG than those older binaries think is valid.
Fix this by having newfs/makefs artificially restrict the number of blocks
in a CG to the slightly smaller value that those older binaries will accept.
The validation code will continue to accept the slightly larger value
that the current newfs/makefs (before this change) could create.
Fixes: 0a6e34e950cd5889122a199c34519b67569be9cc
Reviewed by: mckusick
MFC after: 3 days
Sponsored by: Netflix
fbt/x86: update FBT_AFRAMES to match the trap handlers
Reviewed by: markj
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D40054
fsck_ffs: fix the previous change that skipped pass 5 in some cases
The previous change involved calling check_cgmagic() twice in a row
for the same CG in order to differentiate when the CG was already ok vs.
when the CG was rebuilt, but that doesn't work because the second call
(which was supposed to rebuild the CG) returns 0 (indicating that
the CG was not rebuilt) due to the prevfailcg check causing an early
failure return. Fix this by moving the rebuild part of check_cgmagic()
out into a separate function which is called by pass1() when it wants to
rebuild a CG.
Fixes: da86e7a20dc4a4b17e8d9e7630ed9b675cf71702
Reported by: pho
Discussed with: mckusick
Sponsored by: Netflix
(cherry picked from commit 406475581246360faa7af9891dc8ad2423284e18)
fsck_ffs: fix the previous change that skipped pass 5 in some cases
The previous change involved calling check_cgmagic() twice in a row
for the same CG in order to differentiate when the CG was already ok vs.
when the CG was rebuilt, but that doesn't work because the second call
(which was supposed to rebuild the CG) returns 0 (indicating that
the CG was not rebuilt) due to the prevfailcg check causing an early
failure return. Fix this by moving the rebuild part of check_cgmagic()
out into a separate function which is called by pass1() when it wants to
rebuild a CG.
Fixes: da86e7a20dc4a4b17e8d9e7630ed9b675cf71702
Reported by: pho
Discussed with: mckusick
Sponsored by: Netflix
restore: fix restore of NFS4 ACLs
Changing the mode bits on a file with an NFS4 ACL results in the
NFS4 ACL being replaced by one matching the new mode bits being set,
so when restoring a file with an NFS4 ACL, set the owner/group/mode first
and then set the NFS4 ACL, so that setting the mode does not throw away
the ACL that we just set.
Reviewed by: mckusick
Differential Revision: https://reviews.freebsd.org/D37618
ipmi: use a queue for kcs driver requests when possible
The ipmi watchdog pretimeout action can trigger unintentionally in
certain rare, complicated situations. What we have seen at Netflix
is that the BMC can sometimes be sent a continuous stream of
writes to port 0x80, and due to what is a bug or misconfiguration
in the BMC software, this results in the BMC running out of memory,
becoming very slow to respond to KCS requests, and eventually being
rebooted by its own internal watchdog. While that is going on in
the BMC, back in the host OS, a number of requests are pending in
the ipmi request queue, and the kcs_loop thread is working on
processing these requests. All of the KCS accesses to process
those requests are timing out and eventually failing because the
BMC is responding very slowly or not at all, and the kcs_loop thread
is holding the IPMI_IO_LOCK the whole time that is going on.
Meanwhile the watchdogd process in the host is trying to pat the
BMC watchdog, and this process is sleeping waiting to get the
IPMI_IO_LOCK. It's not entirely clear why the watchdogd process
is sleeping for this lock, because the intention is that a thread
[33 lines not shown]ffs: fix vn_read_from_obj() usage for PAGE_SIZE > block size
vn_read_from_obj() requires that all pages of a vnode (except the last
partial page) be either completely valid or completely invalid,
but for file systems with block size smaller than PAGE_SIZE,
partially valid pages may exist anywhere in the file.
Do not enable the vn_read_from_obj() path in this case.
Reviewed by: mckusick, kib, markj
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D34836
vfs: fix vfs_bio_clrbuf() for PAGE_SIZE > block size
Calculate the desired page valid mask using math that will not
overflow the types used.
Sponsored by: Netflix
Reviewed by: mckusick, kib, markj
Differential Revision: https://reviews.freebsd.org/D34837
ffs: fix vn_io_fault_pgmove() offset for PAGE_SIZE > block size
The "offset" argument to vn_io_fault_pgmove() is supposed to be
the offset within the page, but for ffs we currently use the offset
within the block. When the block size is at least as large as the
page size then these values are the same, but when the page size is
larger than the block size then we need to add the offset of
the block within the page as well.
Sponsored by: Netflix
Reviewed by: mckusick, kib, markj
Differential Revision: https://reviews.freebsd.org/D34835
ffs: wait for trims earlier during unmount to avoid panic
All softdep processing is supposed to be completed by
softdep_flushfiles() and no more deps are supposed to be created after
that, but if a pending trim completes after softdep_flushfiles() and
before softdep_unmount() then the blkfree that is performed by
ffs_blkfree_trim_task() will create a dep when none should exist, and
if softdep_unmount() is called before that dep is freed then the
kernel will panic. Prevent this by waiting for trims to complete
earlier in the unmount process, in ffs_flushfiles(), so that any deps
will be freed and any modified CG buffers will be flushed by the final
fsync of the devvp in ffs_flushfiles() as intended.
Reviewed by: mckusick, kib
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D34806
dtrace: remove unnecessary fflush()
This call was added back in the early days of dtrace porting and
no one knows why anymore. The extra flushing causes lots of
unnecessary CPU overhead when a script produces lots of output,
as well as easily losing output because the command can't keep up.
Sponsored by: Netflix
Reviewed by: imp, markj
Differential Revision: https://reviews.freebsd.org/D34216
librtld_db: Handle shlibs with discontiguous mappings.
Some shared libraries specify mappings that leave a gap (actually a
MAP_GUARD mapping) in between the file mappings (libcrypto.so was
the one I found), and this would cause rd_loadobj_iter() to report
the mapping info incorrectly, leaving out rdl_path and misreporting
rdl_offset for file mappings after the gap. Fix rd_loadobj_iter()
to handle this situation.
Reviewed by: markj
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D32950
fsck_ffs: don't try to write in read-only mode
Skip trying to change fs_mtime for SU+J if we are running read-only.
Reviewed by: mckusick
Sponsored by: Netflix
fsdb: add missing bufinit() call
The bufinit() call in fsck_ffs was moved in commit f190f9193bc10
from a function that is shared with fsdb to one that is private to fsck_ffs,
so add a bufinit() call in fsdb to compensate for that.
Reviewed by: mckusick
Sponsored by: Netflix
tail: fix "tail -F" file rotation detection
When checking if the newly opened file is the same as the old one,
we need to fstat() the new file descriptor, not the old one again.
Reviewed by: glebius
Sponsored by: Netflix
(cherry picked from commit 7787e7eed9d2a43fb6eb66666040f1b495995a2f)
tail: fix "tail -F" file rotation detection
When checking if the newly opened file is the same as the old one,
we need to fstat() the new file descriptor, not the old one again.
Reviewed by: glebius
Sponsored by: Netflix
vfs: fix rangelock range in vn_rdwr() for IO_APPEND
vn_rdwr() must lock the entire file range for IO_APPEND
just like vn_io_fault() does for O_APPEND.
Reviewed by: kib, imp, mckusick
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D28008
ufs: restore uniqueness of st_dev as returned by ufs_stat()
switch ufs_stat() to use the same value for st_dev as was used by
the previous ufs_getattr() stat path.
Submitted by: gallatin
Reviewed by: mjg, imp, kib, mckusick
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D26596
vfs: avoid exposing partially constructed vnodes
If multiple threads race calling vfs_hash_insert() while creating vnodes
with the same identity, all of the vnodes which lose the race must be
destroyed before any other thread can see them. Previously this was
accomplished by the vput() in vfs_hash_insert() resulting in the vnode's
VOP_INACTIVE() method calling vgone() before the vnode lock was unlocked,
but at some point changes to the the vnode refcount/inactive logic have caused
that to no longer work, leading to crashes, so instead vfs_hash_insert()
must call vgone() itself before calling vput() on vnodes which lose the race.
Reviewed by: mjg, kib
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D26291
