CheckIORange is analagous to Linux's access_ok() method, which is checked when
copying in IOVecs in both lib/iov_iter.c:import_single_range() and
lib/iov_iter.c:import_iovec() => fs/read_write.c:rw_copy_check_uvector().
gVisor copies in IOVecs via Task.SingleIOSequence() and Task.CopyInIovecs().
We were checking the address range bounds, but not whether the address is
valid. To conform with linux, we should also check that the address is valid.
For usual preadv/pwritev syscalls, the effect of this change is not noticeable,
since we find out that the address is invalid before the syscall completes.
For vectorized async-IO operations, however, this change is necessary because
Linux returns EFAULT when the operation is submitted, but before it executes.
Thus, we must validate the iovecs when copying them in.
PiperOrigin-RevId: 202370092
Change-Id: I8759a63ccf7e6b90d90d30f78ab8935a0fcf4936
If the child stubs are killed by any unmaskable signal (e.g. SIGKILL), then
the parent process will similarly be killed, resulting in the death of all
other stubs.
The effect of this is that if the OOM killer selects and kills a stub, the
effect is the same as though the OOM killer selected and killed the sentry.
PiperOrigin-RevId: 202219984
Change-Id: I0b638ce7e59e0a0f4d5cde12a7d05242673049d7
IsChrooted still has the opportunity to race with another thread
entering the FSContext into a chroot, but that is unchanged (and
fine, AFAIK).
PiperOrigin-RevId: 202029117
Change-Id: I38bce763b3a7715fa6ae98aa200a19d51a0235f1
The interfaces and their addresses are already available via
the stack Intefaces and InterfaceAddrs.
Also add some tests as we had no tests around SIOCGIFCONF. I also added the socket_netgofer lifecycle for IOCTL tests.
PiperOrigin-RevId: 201744863
Change-Id: Ie0a285a2a2f859fa0cafada13201d5941b95499a
The shutdown behavior where we return EAGAIN for sockets
which are non-blocking is only correct for packet based sockets.
SOCK_STREAM sockets should return EOF.
PiperOrigin-RevId: 201703055
Change-Id: I20b25ceca7286c37766936475855959706fc5397
SendExternalSignal is no longer called before CreateProcess, so it can
enforce this simplified precondition.
StartForwarding, and after Kernel.Start.
PiperOrigin-RevId: 201591170
Change-Id: Ib7022ef7895612d7d82a00942ab59fa433c4d6e9
SIGUSR2 was being masked out to be used as a way to dump sentry
stacks. This could cause compatibility problems in cases anyone
uses SIGUSR2 to communicate with the container init process.
PiperOrigin-RevId: 201575374
Change-Id: I312246e828f38ad059139bb45b8addc2ed055d74
FIOASYNC and friends are used to send signals when a file is ready for IO.
This may or may not be needed by Nginx. While Nginx does use it, it is unclear
if the code that uses it has any effect.
PiperOrigin-RevId: 201550828
Change-Id: I7ba05a7db4eb2dfffde11e9bd9a35b65b98d7f50
Almost all of the hundreds of pending signal queues are empty upon save.
PiperOrigin-RevId: 201380318
Change-Id: I40747072435299de681d646e0862efac0637e172
After shutdown(SHUT_RD) calls to recv /w MSG_DONTWAIT or with
O_NONBLOCK should result in a EAGAIN and not 0. Blocking sockets
should return 0 as they would have otherwise blocked indefinitely.
PiperOrigin-RevId: 201271123
Change-Id: If589b69c17fa5b9ff05bcf9e44024da9588c8876
Instead, CPUs will be created dynamically. We also allow a relatively
efficient mechanism for stealing and notifying when a vCPU becomes
available via unlock.
Since the number of vCPUs is no longer fixed at machine creation time,
we make the dirtySet packing more efficient. This has the pleasant side
effect of cutting out the unsafe address space code.
PiperOrigin-RevId: 201266691
Change-Id: I275c73525a4f38e3714b9ac0fd88731c26adfe66
Resume checks the status of the container and unpauses the kernel
if its status is paused. Otherwise nothing happens.
Tests were added to ensure that the process is in the correct state
after various commands.
PiperOrigin-RevId: 201251234
Change-Id: Ifd11b336c33b654fea6238738f864fcf2bf81e19
Correct a data race in rpcinet where a shutdown and recvmsg can
race around shutown flags.
PiperOrigin-RevId: 201238366
Change-Id: I5eb06df4a2b4eba331eeb5de19076213081d581f
The new policy is identical to FSCACHE (which caches everything in memory), but
it also flushes writes to the backing fs agent immediately.
All gofer cache policy decisions have been moved into the cachePolicy type.
Previously they were sprinkled around the codebase.
There are many different things that we cache (page cache, negative dirents,
dirent LRU, unstable attrs, readdir results....), and I don't think we should
have individual flags to control each of these. Instead, we should have a few
high-level cache policies that are consistent and useful to users. This
refactoring makes it easy to add more such policies.
PiperOrigin-RevId: 201206937
Change-Id: I6e225c382b2e5e1b0ad4ccf8ca229873f4cd389d
Because rpcinet will emulate a blocking socket backed by an rpc based
non-blocking socket. In the event of a shutdown(SHUT_RD) followed by a
read a non-blocking socket is allowed to return an EWOULDBLOCK however
since a blocking socket knows it cannot receive anymore data it would
block indefinitely and in this situation linux returns 0. We have to
track this on the rpcinet sentry side so we can emulate that behavior
because the remote side has no way to know if the socket is actually
blocking within the sentry.
PiperOrigin-RevId: 201201618
Change-Id: I4ac3a7b74b5dae471ab97c2e7d33b83f425aedac
Add support for control messages, but at this time the only
control message that the sentry will support here is SO_TIMESTAMP.
PiperOrigin-RevId: 200922230
Change-Id: I63a852d9305255625d9df1d989bd46a66e93c446
There are circumstances under which the redpill call will not generate
the appropriate action and notification. Replace this call with an
explicit notification, which is guaranteed to transition as well as
perform the futex wake.
PiperOrigin-RevId: 200726934
Change-Id: Ie19e008a6007692dd7335a31a8b59f0af6e54aaa
Boot loader tries to stat mount to determine whether it's a file or not. This
may file if the sandbox process doesn't have access to the file. Instead, add
overlay on top of file, which is better anyway since we don't want to propagate
changes to the host.
PiperOrigin-RevId: 200411261
Change-Id: I14222410e8bc00ed037b779a1883d503843ffebb
Rpcinet already inherits socket.ReceiveTimeout; however, it's
never set on setsockopt(2). The value is currently forwarded
as an RPC and ignored as all sockets will be non-blocking
on the RPC side.
PiperOrigin-RevId: 200299260
Change-Id: I6c610ea22c808ff6420c63759dccfaeab17959dd
This is the first iteration of checkpoint that actually saves to a file.
Tests for checkpoint are included.
Ran into an issue when private unix sockets are enabled. An error message
was added for this case and the mutex state was set.
PiperOrigin-RevId: 200269470
Change-Id: I28d29a9f92c44bf73dc4a4b12ae0509ee4070e93
In order to minimize the likelihood of exit during page table
modifications, make the full set of page table functions split-safe.
This is not strictly necessary (and you may still incur splits due to
allocations from the allocator pool) but should make retries a very rare
occurance.
PiperOrigin-RevId: 200146688
Change-Id: I8fa36aa16b807beda2f0b057be60038258e8d597
hostinet/socket.go: the Sentry doesn't spawn new processes, but it doesn't hurt to protect the socket from leaking.
unet/unet.go: should be setting closing on exec. The FD is explicitly donated to children when needed.
PiperOrigin-RevId: 200135682
Change-Id: Ia8a45ced1e00a19420c8611b12e7a8ee770f89cb
SOCK_STREAM has special behavior with respect to MSG_TRUNC. Specifically,
the data isn't actually copied back out to userspace when MSG_TRUNC is
provided on a SOCK_STREAM.
According to tcp(7): "Since version 2.4, Linux supports the use of
MSG_TRUNC in the flags argument of recv(2) (and recvmsg(2)). This flag
causes the received bytes of data to be discarded, rather than passed
back in a caller-supplied buffer."
PiperOrigin-RevId: 200134860
Change-Id: I70f17a5f60ffe7794c3f0cfafd131c069202e90d
Minor refactor. line_discipline.go was home to 2 large structs (lineDiscipline
and queue), and queue is now large enough IMO to get its own file.
Also moves queue locks into the queue struct, making locking simpler.
PiperOrigin-RevId: 200080301
Change-Id: Ia75a0e9b3d9ac8d7e5a0f0099a54e1f5b8bdea34
Walking off the bottom of the sigaltstack, for example with recursive faults,
results in forced signal delivery, not resetting the stack or pushing signal
stack to whatever happens to lie below the signal stack.
PiperOrigin-RevId: 199856085
Change-Id: I0004d2523f0df35d18714de2685b3eaa147837e0
MSG_TRUNC can cause recvmsg(2) to return a value larger than
the buffer size. In this situation it's an indication that the
buffer was completely filled and that the msg was truncated.
Previously in rpcinet we were returning the buffer size but we
should actually be returning the payload length as returned by
the syscall.
PiperOrigin-RevId: 199814221
Change-Id: If09aa364219c1bf193603896fcc0dc5c55e85d21
Adds support for echo to terminals. Echoing is just copying input back out to
the user, e.g. when I type "foo" into a terminal, I expect "foo" to be echoed
back to my terminal.
Also makes the transform function part of the queue, eliminating the need to
pass them around together and the possibility of using the wrong transform for a
queue.
PiperOrigin-RevId: 199655147
Change-Id: I37c490d4fc1ee91da20ae58ba1f884a5c14fd0d8
Because of the KVM shadow page table implementation, modifications made
to guest page tables from host mode may not be syncronized correctly,
resulting in undefined behavior. This is a KVM bug: page table pages
should also be tracked for host modifications and resynced appropriately
(e.g. the guest could "DMA" into a page table page in theory).
However, since we can't rely on this being fixed everywhere, workaround
the issue by forcing page table modifications to be in guest mode. This
will generally be the case anyways, but now if an exit occurs during
modifications, we will re-enter and perform the modifications again.
PiperOrigin-RevId: 199587895
Change-Id: I83c20b4cf2a9f9fa56f59f34939601dd34538fb0
Instead of associating a single PCID with each set of page tables (which
will reach the maximum quickly), allow a dynamic pool for each vCPU.
This is the same way that Linux operates. We also split management of
PCIDs out of the page tables themselves for simplicity.
PiperOrigin-RevId: 199585631
Change-Id: I42f3486ada3cb2a26f623c65ac279b473ae63201
In order to prevent possible garbage collection and reuse of page table
pages prior to invalidation, introduce a former allocator abstraction
that can ensure entries are held during a single traversal. This also
cleans up the abstraction and splits it out of the machine itself.
PiperOrigin-RevId: 199581636
Change-Id: I2257d5d7ffd9c36f9b7ecd42f769261baeaf115c
This change will add support for ioctls that have previously
been supported by netstack.
LINE_LENGTH_IGNORE
PiperOrigin-RevId: 199544114
Change-Id: I3769202c19502c3b7d05e06ea9552acfd9255893
This change will add support for /proc/sys/net and /proc/net which will
be managed and owned by rpcinet. This will allow these inodes to be forward
as rpcs.
PiperOrigin-RevId: 199370799
Change-Id: I2c876005d98fe55dd126145163bee5a645458ce4
This is necessary to prevent races with invalidation. It is currently
possible that page tables are garbage collected while paging caches
refer to them. We must ensure that pages are held until caches can be
invalidated. This is not achieved by this goal alone, but moving locking
to outside the page tables themselves is a requisite.
PiperOrigin-RevId: 198920784
Change-Id: I66fffecd49cb14aa2e676a84a68cabfc0c8b3e9a
Previously, the vCPU FS was always correct because it relied on the
reset coming out of the switch. When that doesn't occur, for example,
using bluepill directly, the FS value can be incorrect leading to
strange corruption.
This change is necessary for a subsequent change that enforces guest
mode for page table modifications, and it may reduce test flakiness.
(The problematic path may occur in tests, but does not occur in the
actual platform.)
PiperOrigin-RevId: 198648137
Change-Id: I513910a973dd8666c9a1d18cf78990964d6a644d
This is a refactor of ring0 and ring0/pagetables that changes from
individual arguments to opts structures. This should involve no
functional changes, but sets the stage for subsequent changes.
PiperOrigin-RevId: 198627556
Change-Id: Id4460340f6a73f0c793cd879324398139cd58ae9
These were causing non-blocking related errnos to be returned to
the sentry when they were created as blocking FDs internally.
PiperOrigin-RevId: 197962932
Change-Id: I3f843535ff87ebf4cb5827e9f3d26abfb79461b0
Kernel before 2.6.16 return EINVAL, but later return ESPIPE for this case.
Also change type of "length" from Uint(uint32) to Int64.
Because C header uses type "size_t" (unsigned long) or "off_t" (long) for length.
And it makes more sense to check length < 0 with Int64 because Uint cannot be negative.
Change-Id: Ifd7fea2dcded7577a30760558d0d31f479f074c4
PiperOrigin-RevId: 197616743
Establishes a way of communicating interface flags between netstack and
epsocket. More flags can be added over time.
PiperOrigin-RevId: 197616669
Change-Id: I230448c5fb5b7d2e8d69b41a451eb4e1096a0e30
Especially in situations with small numbers of vCPUs, the existing
system resulted in excessive thrashing. Now, execution contexts
co-ordinate as smoothly as they can to share a small number of cores.
PiperOrigin-RevId: 197483323
Change-Id: I0afc0c5363ea9386994355baf3904bf5fe08c56c
In Linux, many UDS ioctls are passed through to the NIC driver. We do the same
here, passing ioctl calls to Unix sockets through to epsocket.
In Linux you can see this path at net/socket.c:sock_ioctl, which calls
sock_do_ioctl, which calls net/core/dev_ioctl.c:dev_ioctl.
SIOCGIFNAME is also added.
PiperOrigin-RevId: 197167508
Change-Id: I62c326a4792bd0a473e9c9108aafb6a6354f2b64
Capabilities for sysv sem operations were being checked against the
current task's user namespace. They should be checked against the user
namespace owning the ipc namespace for the sems instead, per
ipc/util.c:ipcperms().
PiperOrigin-RevId: 197063111
Change-Id: Iba29486b316f2e01ee331dda4e48a6ab7960d589
This should fix the socket Dirent memory leak.
fs.NewFile takes a new reference. It should hold the *only* reference.
DecRef that socket Dirent.
Before the globalDirentMap was introduced, a mis-refcounted Dirent
would be garbage collected when all references to it were gone. For
socket Dirents, this meant that they would be garbage collected when
the associated fs.Files disappeared.
After the globalDirentMap, Dirents *must* be reference-counted
correctly to be garbage collected, as Dirents remove themselves
from the global map when their refcount goes to -1 (see Dirent.destroy).
That removes the last pointer to that Dirent.
PiperOrigin-RevId: 196878973
Change-Id: Ic7afcd1de97c7101ccb13be5fc31de0fb50963f0
When doing a BidirectionalConnect we don't need to continue holding
the ConnectingEndpoint's mutex when creating the NewConnectedEndpoint
as it was held during the Connect. Additionally, we're not holding
the baseEndpoint mutex while Unregistering an event.
PiperOrigin-RevId: 196875557
Change-Id: Ied4ceed89de883121c6cba81bc62aa3a8549b1e9
Previously, inet.Stack was referenced in 2 structs in sentry/socket that can be
saved/restored. If an app is saved and restored on another machine, it may try
to use the old stack, which will have been replaced by a new stack on the new
machine.
PiperOrigin-RevId: 196733985
Change-Id: I6a8cfe73b5d7a90749734677dada635ab3389cb9
When the amount of data read is more than the amount written, sendfile would not
adjust 'in file' position and would resume from the wrong location.
Closes#33
PiperOrigin-RevId: 196731287
Change-Id: Ia219895dd765016ed9e571fd5b366963c99afb27
When file is backed by host FD, atime and mtime for the host file and the
cached attributes in the Sentry must be close together. In this case,
the call to update atime and mtime can be skipped. This is important when
host filesystem is using overlay because updating atime and mtime explicitly
forces a copy up for every file that is touched.
PiperOrigin-RevId: 196176413
Change-Id: I3933ea91637a071ba2ea9db9d8ac7cdba5dc0482
Otherwise, mounts that fail to be unmounted (EBUSY) will be removed
from the children list anyway.
At this point, this just affects /proc/pid/mounts and /proc/pid/mountinfo.
PiperOrigin-RevId: 195267588
Change-Id: I79114483d73b90f9a7d764a7d513b5b2f251182e
Detachable exec commands are handled in the client entirely and the detach option is not used anymore.
PiperOrigin-RevId: 195181272
Change-Id: I6e82a2876d2c173709c099be59670f71702e5bf0
As of Linux 4.15 (f29810335965ac1f7bcb501ee2af5f039f792416
KVM/x86: Check input paging mode when cs.l is set), KVM validates that
LMA is set along with LME.
PiperOrigin-RevId: 195047401
Change-Id: I8b43d8f758a85b1f58ccbd747dcacd4056ef3f66
Nicolas Lacasse
Fabricio Voznika
Jamie Liu
Adin Scannell
Brian Geffon
Michael Pratt
Zhaozhong Ni
Ian Gudger
Justine Olshan
Brielle Broder
Kevin Krakauer
Googler
Chanwit Kaewkasi
Rahat Mahmood
Christopher Koch
Cyrille Hemidy
Zhengyu He