John Baldwin [Tue, 5 May 2020 00:02:04 +0000 (00:02 +0000)]
Initial support for bhyve save and restore.
Save and restore (also known as suspend and resume) permits a snapshot
to be taken of a guest's state that can later be resumed. In the
current implementation, bhyve(8) creates a UNIX domain socket that is
used by bhyvectl(8) to send a request to save a snapshot (and
optionally exit after the snapshot has been taken). A snapshot
currently consists of two files: the first holds a copy of guest RAM,
and the second file holds other guest state such as vCPU register
values and device model state.
To resume a guest, bhyve(8) must be started with a matching pair of
command line arguments to instantiate the same set of device models as
well as a pointer to the saved snapshot.
While the current implementation is useful for several uses cases, it
has a few limitations. The file format for saving the guest state is
tied to the ABI of internal bhyve structures and is not
self-describing (in that it does not communicate the set of device
models present in the system). In addition, the state saved for some
device models closely matches the internal data structures which might
prove a challenge for compatibility of snapshot files across a range
of bhyve versions. The file format also does not currently support
versioning of individual chunks of state. As a result, the current
file format is not a fixed binary format and future revisions to save
and restore will break binary compatiblity of snapshot files. The
goal is to move to a more flexible format that adds versioning,
etc. and at that point to commit to providing a reasonable level of
compatibility. As a result, the current implementation is not enabled
by default. It can be enabled via the WITH_BHYVE_SNAPSHOT=yes option
for userland builds, and the kernel option BHYVE_SHAPSHOT.
Submitted by: Mihai Tiganus, Flavius Anton, Darius Mihai
Submitted by: Elena Mihailescu, Mihai Carabas, Sergiu Weisz
Relnotes: yes
Sponsored by: University Politehnica of Bucharest
Sponsored by: Matthew Grooms (student scholarships)
Sponsored by: iXsystems
Differential Revision: https://reviews.freebsd.org/D19495
Randall Stewart [Mon, 4 May 2020 23:02:58 +0000 (23:02 +0000)]
This fixes two issues found by ankitraheja09@gmail.com
1) When BBR retransmits the syn it was messing up the snd_max
2) When we need to send a RST we might not send it when we should
Michael Tuexen [Mon, 4 May 2020 22:02:49 +0000 (22:02 +0000)]
Enter the net epoch before calling the output routine in TCP BBR.
This was only triggered when setting the IPPROTO_TCP level socket
option TCP_DELACK.
This issue was found by runnning an instance of SYZKALLER.
Reviewed by: rrs
Sponsored by: Netflix, Inc.
Differential Revision: https://reviews.freebsd.org/D24690
Randall Stewart [Mon, 4 May 2020 20:28:53 +0000 (20:28 +0000)]
This commit brings things into sync with the advancements that
have been made in rack and adds a few fixes in BBR. This also
removes any possibility of incorrectly doing OOB data the stacks
do not support it. Should fix the skyzaller crashes seen in the
past. Still to fix is the BBR issue just reported this weekend
with the SYN and on sending a RST. Note that this version of
rack can now do pacing as well.
Randall Stewart [Mon, 4 May 2020 20:19:57 +0000 (20:19 +0000)]
Adjust the fb to have a way to ask the underlying stack
if it can support the PRUS option (OOB). And then have
the new function call that to validate and give the
correct error response if needed to the user (rack
and bbr do not support obsoleted OOB data).
Sponsoered by: Netflix Inc.
Differential Revision: https://reviews.freebsd.org/D24574
Brooks Davis [Mon, 4 May 2020 17:16:30 +0000 (17:16 +0000)]
Set LG_VADDR to 48 on RISC-V.
The Sv48 PTE format is the largest currently defined address space for
RISC-V. It makes no sense to define a larger size and doing so (at
least for 64-bits) forces rtrees down a slow path.
After converting routing subsystem customers to use nexthop objects
defined in r359823, some fields in struct rtentry became unused.
This commit removes rt_ifp, rt_ifa, rt_gateway and rt_mtu from struct rtentry
along with the code initializing and updating these fields.
Cleanup of the remaining fields will be addressed by D24669.
This commit also changes the implementation of the RTM_CHANGE handling.
Old implementation tried to perform the whole operation under radix WLOCK,
resulting in slow performance and hacks like using RTF_RNH_LOCKED flag.
New implementation looks up the route nexthop under radix RLOCK, creates new
nexthop and tries to update rte nhop pointer. Only last part is done under
WLOCK.
In the hypothetical scenarious where multiple rtsock clients
repeatedly issue RTM_CHANGE requests for the same route, route may get
updated between read and update operation. This is addressed by retrying
the operation multiple (3) times before returning failure back to the
caller.
[evdev] Add AT translated set1 scancodes for F-unlocked F1-12 keys.
"F lock" is a switch between two sets of scancodes for function keys F1-F12
found on some Logitech and Microsoft PS/2 keyboards [1]. When "F lock" is
pressed, then F1-F12 act as function keys and produce usual keyscans for
these keys. When "F lock" is depressed, F1-F12 produced the same keyscans
but prefixed with E0.
Some laptops use [2] E0-prefixed F1-F12 scancodes for non-standard keys.
Alexander Motin [Sun, 3 May 2020 16:14:55 +0000 (16:14 +0000)]
Add session locking in cfiscsi_ioctl_handoff().
While there, remove ifdef around cs_target check in cfiscsi_ioctl_list().
I am not sure why this ifdef was added, but without this check code will
crash below on NULL dereference.
Fix various, mostly minor errors in man pages like:
- Abbreviated month name in .Dd
- position of HISTORY section
- alphabetical ordering within SEE ALSO section
- adding .Ed before .Sh DESCRIPTION
- remove trailing whitespaces
- Line break after a sentence stop
- Use BSD OS macros instead of hardcoded strings
No .Dd bumps as there was no actual content change made
in any of these pages.
Warner Losh [Sun, 3 May 2020 04:22:27 +0000 (04:22 +0000)]
We need to hold the periph lock when we release the ccb (and when we
run it). Make sure that we do. Simplify the flow a bit, and fix a
comment since we do need to do these things.
Gleb Smirnoff [Sun, 3 May 2020 00:37:16 +0000 (00:37 +0000)]
Step 4.2: start divorce of M_EXT and M_EXTPG
They have more differencies than similarities. For now there is lots
of code that would check for M_EXT only and work correctly on M_EXTPG
buffers, so still carry M_EXT bit together with M_EXTPG. However,
prepare some code for explicit check for M_EXTPG.
Gleb Smirnoff [Sun, 3 May 2020 00:12:56 +0000 (00:12 +0000)]
Step 3: anonymize struct mbuf_ext_pgs and move all its fields into mbuf
within m_epg namespace.
All edits except the 'struct mbuf' declaration and mb_dupcl() were done
mechanically with sed:
Gleb Smirnoff [Sat, 2 May 2020 23:46:29 +0000 (23:46 +0000)]
Step 2.2:
o Shrink sglist(9) functions to work with multipage mbufs down from
four functions to two.
o Don't use 'struct mbuf_ext_pgs *' as argument, use struct mbuf.
o Rename to something matching _epg.
Gleb Smirnoff [Sat, 2 May 2020 22:44:23 +0000 (22:44 +0000)]
In mb_unmapped_compress() we don't need mbuf structure to keep data,
but we need buffer of MLEN bytes. This isn't just a simplification,
but important fixup, because previous commit shrinked sizeof(struct
mbuf) down below MSIZE, and instantiating an mbuf on stack no longer
provides enough data.
Gleb Smirnoff [Sat, 2 May 2020 22:39:26 +0000 (22:39 +0000)]
Continuation of multi page mbuf redesign from r359919.
The following series of patches addresses three things:
Now that array of pages is embedded into mbuf, we no longer need
separate structure to pass around, so struct mbuf_ext_pgs is an
artifact of the first implementation. And struct mbuf_ext_pgs_data
is a crutch to accomodate the main idea r359919 with minimal churn.
Also, M_EXT of type EXT_PGS are just a synonym of M_NOMAP.
The namespace for the newfeature is somewhat inconsistent and
sometimes has a lengthy prefixes. In these patches we will
gradually bring the namespace to "m_epg" prefix for all mbuf
fields and most functions.
Step 1 of 4:
o Anonymize mbuf_ext_pgs_data, embed in m_ext
o Embed mbuf_ext_pgs
o Start documenting all this entanglement
Alan Somers [Sat, 2 May 2020 20:14:59 +0000 (20:14 +0000)]
Resolve conflict between the fusefs(5) and mac_bsdextended(4) tests
mac_bsdextended(4), when enabled, causes ordinary operations to send many
more VOP_GETATTRs to file system. The fusefs tests expectations aren't
written with those in mind. Optionally expecting them would greatly
obfuscate the fusefs tests. Worse, certain fusefs functionality (like
attribute caching) would be impossible to test if the tests couldn't expect
an exact number of GETATTR operations.
This commit resolves that conflict by making two changes:
1. The fusefs tests will now check for mac_bsdextended, and skip if it's
enabled.
2. The mac_bsdextended tests will now check whether the module is enabled, not
merely loaded. If it's loaded but disabled, the tests will automatically
enable it for the duration of the tests.
With these changes, a CI system can achieve best coverage by loading both
fusefs and mac_bsdextended at boot, and setting
security.mac.bsdextended.enabled=0
Alexander Motin [Sat, 2 May 2020 16:54:59 +0000 (16:54 +0000)]
Cleanup LUN addition/removal.
- Make ctl_add_lun() synchronous. Asynchronous addition was used by
Copan's proprietary code long ago and never for upstream FreeBSD.
- Move LUN enable/disable calls from backends to CTL core.
- Serialize LUN modification and partially removal to avoid double frees.
- Slightly unify backends code.
John Baldwin [Sat, 2 May 2020 01:00:29 +0000 (01:00 +0000)]
Don't pass bogus keys down for NULL algorithms.
The changes in r359374 added various sanity checks in sessions and
requests created by crypto consumers in part to permit backend drivers
to make assumptions instead of duplicating checks for various edge
cases. One of the new checks was to reject sessions which provide a
pointer to a key while claiming the key is zero bits long.
IPsec ESP tripped over this as it passes along whatever key is
provided for NULL, including a pointer to a zero-length key when an
empty string ("") is used with setkey(8). One option would be to
teach the IPsec key layer to not allocate keys of zero length, but I
went with a simpler fix of just not passing any keys down and always
using a key length of zero for NULL algorithms.
John Baldwin [Sat, 2 May 2020 00:06:58 +0000 (00:06 +0000)]
Remove support for IPsec algorithms deprecated in r348205 and r360202.
Examples of depecrated algorithms in manual pages and sample configs
are updated where relevant. I removed the one example of combining
ESP and AH (vs using a cipher and auth in ESP) as RFC 8221 says this
combination is NOT RECOMMENDED.
Specifically, this removes support for the following ciphers:
- des-cbc
- 3des-cbc
- blowfish-cbc
- cast128-cbc
- des-deriv
- des-32iv
- camellia-cbc
This also removes support for the following authentication algorithms:
- hmac-md5
- keyed-md5
- keyed-sha1
- hmac-ripemd160
Mitchell Horne [Fri, 1 May 2020 21:58:19 +0000 (21:58 +0000)]
Use the HSM SBI extension to start APs
The addition of the HSM SBI extension to OpenSBI introduces a new
breaking change: secondary harts will remain parked in the firmware,
until they are brought up explicitly via sbi_hsm_hart_start(). Add
the call to do this, sending the secondary harts to mpentry.
If the HSM extension is not present, secondary harts are assumed to be
released by the firmware, as is the case for OpenSBI =< v0.6 and BBL.
In the case that the HSM call fails we exclude the CPU, notify the
user, and allow the system to proceed with booting.
Mitchell Horne [Fri, 1 May 2020 21:52:29 +0000 (21:52 +0000)]
Make mpentry independent of _start
APs enter the kernel at the same point as the BSP, the _start routine.
They then jump to mpentry, but not before storing the kernel's physical
load address in the s9 register. Extract this calculation into its own
routine, so that APs can be instructed to enter directly from mpentry.
Warner Losh [Fri, 1 May 2020 21:24:19 +0000 (21:24 +0000)]
Add KASSERT to ensure sane nsid.
All callers are currently filtering bad nsid to this function,
however, we'll have undefined behavior if that's not true. Add the
KASSERT to prevent that.
Warner Losh [Fri, 1 May 2020 20:29:46 +0000 (20:29 +0000)]
Various improvements to this man page:
o Be consistent about device-id and namespace-id
o Use consistent arg markup for these
o document you can use disk names too
o document nsid command better
o document the idenntify command
o add a couple of examples.
Warner Losh [Fri, 1 May 2020 17:50:21 +0000 (17:50 +0000)]
Remove more stray sparc64 ifdefs.
Also, dmabuf appears to only be set for sparc64 case, but there was a
comment at its only use that says it was broken for some apple
adapters. #ifdef it all of that out now that nothing sets it.
Warner Losh [Fri, 1 May 2020 17:16:57 +0000 (17:16 +0000)]
When we have an invalid build option, don't rm -rf the current
directory.
Add a quick sanity check to objdir before using it. It must start
with /. If there was a make error getting it, report that and continue
with the next target. If there was anything else, bail out.
Brandon Bergren [Fri, 1 May 2020 16:56:36 +0000 (16:56 +0000)]
[PowerPC] Set fixed boot1.elf load address
Due to the way claiming works, we need to ensure on AIM OFW machines that
we don't have overlapping ranges on any step of the load.
Load boot1.elf at 0x38000 so it will not overlap with anything even if the
entire PReP partition gets loaded by OFW.
Tested on an iBook G4, a PowerBook G4, a PowerMac G5, and qemu pseries.
(qemu pseries is broken without this patch due to the high address used
by lld10.)
Previously procctl(PROC_PROTMAX_STATUS, ... used the PROC_ASLR_NOFORCE
macro for the "system-wide configured policy" status, instead of
PROC_PROTMAX_NOFORCE.
They both have a value of 3, so no functional change.
Fix a number of the following issues in man4:
- Inconsistencies in .Dd like abbreviated month names,
"th" after numbers, or leading zeros
- No line breaks after a sentence stop
- Whitespace at the end of the line
- Use macros for BSD OS names instead of hardcoded names
- CAVEATS instead of CAVEAT in section name
No actual content change in terms of additions were made, so
no bump of the .Dd for these man pages.
All of these issues were found and fixed by Gordon Bergling.
Rick Macklem [Fri, 1 May 2020 00:36:14 +0000 (00:36 +0000)]
Factor some code out of nfsm_dissct() into separate functions.
Factoring some of the code in nfsm_dissct() out into separate functions
allows these functions to be used elsewhere in the NFS mbuf handling code.
Other uses of these functions will be done in future commits.
It also makes it easier to add support for ext_pgs mbufs, which is needed
for nfs-over-tls under development in base/projects/nfs-over-tls.
Although the algorithm in nfsm_dissct() is somewhat re-written by this
patch, the semantics of nfsm_dissct() should not have changed.
John Baldwin [Thu, 30 Apr 2020 22:08:40 +0000 (22:08 +0000)]
Remove the SYMVER build option.
This option was added as a transition aide when symbol versioning was
first added. It was enabled by default in 2007 and is supported even
by the old GPLv2 binutils. Trying to disable it currently fails to
build in libc and at this point it isn't worth fixing the build.
Reported by: Michael Dexter
Reviewed by: imp
Differential Revision: https://reviews.freebsd.org/D24637
Warner Losh [Thu, 30 Apr 2020 21:08:59 +0000 (21:08 +0000)]
Fix three bugs highlighted by review:
- maxio should be dp->d_maxsize. This is often MAXPHYS, but not always
(especially if MAXPHYS is > 1MB).
- Unlock the periph before returning. We don't need to relock it to
release the ccb.
- Make sure we release the ccb in error paths.
lualoader: config: improve readConfFiles, rename to readConf
The previous interface was pretty bad, and required the caller to get some
implementation details correct that it really shouldn't need to (e.g.
loader_conf_files handling) and pass in an empty table for it to use.
The new and much improved interface, readConf, is much less of a hack;
hiding these implementation details and just doing the right thing.
config.lua will now use it to process /boot/defaults/loader.conf and the
subsequent loader_conf_files from there, and read-conf will also use it.
This improvement submitted by Olivier (cited below), loader_conf_files
handling from the original patch was changed to just clobber it before
processing and not bother restoring it after the fact following r360505
where it's now guaranteed to evade the loader environment.
lualoader: config: add a table for restricted environment vars
This new table should be used for transient values that don't need to end up
in the loader environment. Generally, these will be things that are internal
details that really aren't needed or interesting outside of the config
module (e.g. if we changed how ${module}_* directives work, they might use
this instead).
To start, populate it with loader_conf_files. Any specific value of
loader_conf_files isn't all that interesting; if we're going to export it,
we should really instead export a loader_conf_files that indicates all of
the configuration files we processed. This will be used to reduce
bookkeeping overhead in a future commit that cleans up readConfFiles.
Ed Maste [Thu, 30 Apr 2020 18:11:56 +0000 (18:11 +0000)]
src.opts.mk: add WITHOUT_OPENSSL dependencies
A number of components require OpenSSL and fail to build if it is not
enabled. As a first phase force these off under WITHOUT_OPENSSL. A
second phase should make these more fine-grained, allowing the component
to build but without OpenSSL.
Running TCP Cubic together with ECN could end up reducing cwnd down to 1 byte, if the
receiver continously sets the ECE flag, resulting in very poor transmission speeds.
In line with RFC6582 App. B, a lower bound of 2 MSS is introduced, as well as a typecast
to prevent any potential integer overflows during intermediate calculation steps of the
adjusted cwnd.
loader.conf(5): document that loader_conf_files may be clobbered
A future change in lualoader may take some liberties with the
loader_conf_files in the name of efficiency; namely, it may start omitting
it from the loader environment entirely so that it doesn't need to worry
about maintaining any specific value.
This variable has historically been incredibly volatile anyways, as it may
get set to completely different values in any given configuration file to
trigger a load of more files.
Document now that we may not maintain it in the future, but perhaps we'll
reserve the right to change our minds and eventually formally export all of
the loader configuration files that were read using this variable.
Warner Losh [Thu, 30 Apr 2020 00:43:07 +0000 (00:43 +0000)]
Implement the NVME_GET_NSID and NVME_PASSTHROUGH_CMD ioctls
With these two ioctls implemented in the nda driver, nvmecontrol now
works with nda just like it does with nvd. It eliminates the need to
jump through odd hoops to get this data.
Warner Losh [Thu, 30 Apr 2020 00:43:02 +0000 (00:43 +0000)]
Return the nvmeX device associated with the ndaX device.
Add the nvmeX device to the XPT_PATH_INQ nvme specific
information. while one could figure this out by looking up the
domain:bus:slot:function, it's a lot easier to have the SIM set it
directly since the sim knows this.
projects / FreeBSD / FreeBSD.git / log