summaryrefslogtreecommitdiffhomepage
AgeCommit message (Collapse)Author
2019-07-02Ext4: DiskLayout: Inode interface.Ayush Ranjan
PiperOrigin-RevId: 256234390
2019-07-02Merge pull request #279 from kevinGC:iptables-1-pkggVisor bot
PiperOrigin-RevId: 256231055
2019-07-02Simplify (and fix) refcounts in createAt.Nicolas Lacasse
fileOpAt holds references on the Dirents passed as arguments to the callback, and drops refs when finished, so we don't need to DecRef those Dirents ourselves However, all Dirents that we get from FindInode/FindLink must be DecRef'd. This CL cleans up the ref-counting logic, and fixes some refcount issues in the process. PiperOrigin-RevId: 256220882
2019-07-02sentry/kernel: add syslog messageAhmet Alp Balkan
It feels like "reticulating splines" is missing from the list of meaningless syslog messages. Signed-off-by: Ahmet Alp Balkan <ahmetb@google.com>
2019-07-01Fix unix/transport.queue reference leaks.Ian Gudger
Fix two leaks for connectionless Unix sockets: * Double connect: Subsequent connects would leak a reference on the previously connected endpoint. * Close unconnected: Sockets which were not connected at the time of closure would leak a reference on their receiver. PiperOrigin-RevId: 256070451
2019-07-01Use new location of python-hello image in tests.Nicolas Lacasse
PiperOrigin-RevId: 256062988
2019-07-01Check remaining traversal limit when creating a file through a symlink.Nicolas Lacasse
This fixes the case when an app tries to create a file that already exists, and is a symlink to itself. A test was added. PiperOrigin-RevId: 256044811
2019-06-29Add stack trace printing to reference leak checking.Ian Gudger
PiperOrigin-RevId: 255759891
2019-06-28Drop local_server support.Adin Scannell
PiperOrigin-RevId: 255713414
2019-06-28Add finalizer on AtomicRefCount to check for leaks.Ian Gudger
PiperOrigin-RevId: 255711454
2019-06-28Drop ashmem and binder.Adin Scannell
These are unfortunately unused and unmaintained. They can be brought back in the future if need requires it. PiperOrigin-RevId: 255697132
2019-06-28Remove events from name_to_handle_at and open_by_handle_at.Nicolas Lacasse
These syscalls require filesystem support that gVisor does not provide, and is not planning to implement. Their absense should not trigger an event. PiperOrigin-RevId: 255692871
2019-06-28ext4: disklayout: SuperBlock interface implementations.Ayush Ranjan
PiperOrigin-RevId: 255687771
2019-06-28Fix suggestions from clang.Nicolas Lacasse
PiperOrigin-RevId: 255679603
2019-06-28Automated rollback of changelist 255263686Nicolas Lacasse
PiperOrigin-RevId: 255679453
2019-06-28platform/ptrace: return more detailed errorsAndrei Vagin
Right now, if we can't create a stub process, we will see this error: panic: unable to activate mm: resource temporarily unavailable It would be better to know the root cause of this "resource temporarily unavailable". PiperOrigin-RevId: 255656831
2019-06-28Superblock interface in the disk layout package for ext4.Ayush Ranjan
PiperOrigin-RevId: 255644277
2019-06-28runsc: allow openat for runsc-raceAndrei Vagin
I see that runsc-race is killed by SIGSYS, because openat isn't allowed by seccomp filters: 60052 openat(AT_FDCWD, "/proc/sys/vm/overcommit_memory", O_RDONLY|O_CLOEXEC <unfinished ...> 60052 <... openat resumed> ) = 257 60052 --- SIGSYS {si_signo=SIGSYS, si_code=SYS_SECCOMP, si_call_addr=0xfaacf1, si_syscall=__NR_openat, si_arch=AUDIT_ARCH_X86_64} --- PiperOrigin-RevId: 255640808
2019-06-28Fix deadloop in proc subtask listYong He
Readdir of /proc/x/task/ will get direntry entries from tasks of specified taskgroup. Now the tasks slice is unsorted, use sort.SearchInts search entry from the slice may cause infinity loops. The fix is sort the slice before search. This issue could be easily reproduced via following steps, revise Readdir in pkg/sentry/fs/proc/task.go, force set taskInts into test slice []int{1, 11, 7, 5, 10, 6, 8, 3, 9, 2, 4}, then run docker image and run ls /proc/1/task, the command will cause infinity loops.
2019-06-27Complete pipe support on overlayfsFabricio Voznika
Get/Set pipe size and ioctl support were missing from overlayfs. It required moving the pipe.Sizer interface to fs so that overlay could get access. Fixes #318 PiperOrigin-RevId: 255511125
2019-06-27Fix various spelling issues in the documentationMichael Pratt
Addresses obvious typos, in the documentation only. COPYBARA_INTEGRATE_REVIEW=https://github.com/google/gvisor/pull/443 from Pixep:fix/documentation-spelling 4d0688164eafaf0b3010e5f4824b35d1e7176d65 PiperOrigin-RevId: 255477779
2019-06-27Cache directory entries in the overlayMichael Pratt
Currently, the overlay dirCache is only used for a single logical use of getdents. i.e., it is discard when the FD is closed or seeked back to the beginning. But the initial work of getting the directory contents can be quite expensive (particularly sorting large directories), so we should keep it as long as possible. This is very similar to the readdirCache in fs/gofer. Since the upper filesystem does not have to allow caching readdir entries, the new CacheReaddir MountSourceOperations method controls this behavior. This caching should be trivially movable to all Inodes if desired, though that adds an additional copy step for non-overlay Inodes. (Overlay Inodes already do the extra copy). PiperOrigin-RevId: 255477592
2019-06-27gvisor/ptrace: grub initial thread registers only onceAndrei Vagin
PiperOrigin-RevId: 255465635
2019-06-27Merge pull request #461 from brb-g:128_procseekendgVisor bot
PiperOrigin-RevId: 255462850
2019-06-26Preserve permissions when checking lowerFabricio Voznika
The code was wrongly assuming that only read access was required from the lower overlay when checking for permissions. This allowed non-writable files to be writable in the overlay. Fixes #316 PiperOrigin-RevId: 255263686
2019-06-26Follow symlinks when creating a file, and create the target.Nicolas Lacasse
If we have a symlink whose target does not exist, creating the symlink (either via 'creat' or 'open' with O_CREAT flag) should create the target of the symlink. Previously, gVisor would error with EEXIST in this case PiperOrigin-RevId: 255232944
2019-06-26Always set SysProcAttr.Ctty to an FD in the child's FD table.Nicolas Lacasse
Go was going to change the behavior of SysProcAttr.Ctty such that it must be an FD in the *parent* FD table: https://go-review.googlesource.com/c/go/+/178919/ However, after some debate, it was decided that this change was too backwards-incompatible, and so it was reverted. https://github.com/golang/go/issues/29458 The behavior going forward is unchanged: the Ctty FD must be an FD in the *child* FD table. PiperOrigin-RevId: 255228476
2019-06-25Add TODO reminder to remove tmpfs caching optionsMichael Pratt
Updates #179 PiperOrigin-RevId: 255081565
2019-06-25Add //pkg/fdchannel.Jamie Liu
To accompany flipcall connections in cases where passing FDs is required (as for gofers). PiperOrigin-RevId: 255062277
2019-06-25Use different Ctty FDs based on the go version.Nicolas Lacasse
An upcoming change in Go 1.13 [1] changes the semantics of the SysProcAttr.Ctty field. Prior to the change, the FD must be an FD in the child process's FD table (aka "post-shuffle"). After the change, the FD must be an FD in the current process's FD table (aka "pre-shuffle"). To be compatible with both versions this CL introduces a new boolean "CttyFdIsPostShuffle" which indicates whether a pre- or post-shuffle FD should be provided. We use build tags to chose the correct one. 1: https://go-review.googlesource.com/c/go/+/178919/ PiperOrigin-RevId: 255015303
2019-06-25gvisor: lockless read access for task credentialsAndrei Vagin
Credentials are immutable and even before these changes we could read them without locks, but we needed to take a task lock to get a credential object from a task object. It is possible to avoid this lock, if we will guarantee that a credential object will not be changed after setting it on a task. PiperOrigin-RevId: 254989492
2019-06-24fsgopher: reopen files via /proc/self/fdAndrei Vagin
When we reopen file by path, we can't be sure that we will open exactly the same file. The file can be deleted and another one with the same name can be created. PiperOrigin-RevId: 254898594
2019-06-24Add CLOCK_BOOTTIME as a CLOCK_MONOTONIC aliasAdrien Leravat
Makes CLOCK_BOOTTIME available with * clock_gettime * timerfd_create * clock_gettime vDSO CLOCK_BOOTTIME is implemented as an alias to CLOCK_MONOTONIC. CLOCK_MONOTONIC already keeps track of time across save and restore. This is the closest possible behavior to Linux CLOCK_BOOTIME, as there is no concept of suspend/resume. Updates google/gvisor#218
2019-06-24fs: synchronize concurrent writes into files with O_APPENDAndrei Vagin
For files with O_APPEND, a file write operation gets a file size and uses it as offset to call an inode write operation. This means that all other operations which can change a file size should be blocked while the write operation doesn't complete. PiperOrigin-RevId: 254873771
2019-06-24Add O_EXITKILL to ptrace options.Adin Scannell
This prevents a race before PDEATH_SIG can take effect during a sentry crash. Discovered and solution by avagin@. PiperOrigin-RevId: 254871534
2019-06-24Implement /proc/net/tcp.Rahat Mahmood
PiperOrigin-RevId: 254854346
2019-06-24platform/ptrace: specify PTRACE_O_TRACEEXIT for stub-processesAndrei Vagin
The tracee is stopped early during process exit, when registers are still available, allowing the tracer to see where the exit occurred, whereas the normal exit notifi? cation is done after the process is finished exiting. Without this option, dumpAndPanic fails to get registers. PiperOrigin-RevId: 254852917
2019-06-24Use correct statx syscall number for amd64.Nicolas Lacasse
The previous number was for the arm architecture. Also change the statx tests to force them to run on gVisor, which would have caught this issue. PiperOrigin-RevId: 254846831
2019-06-24Allow to change logging options using 'runsc debug'Fabricio Voznika
New options are: runsc debug --strace=off|all|function1,function2 runsc debug --log-level=warning|info|debug runsc debug --log-packets=true|false Updates #407 PiperOrigin-RevId: 254843128
2019-06-24Add regression test for #128 (fixed in ab6774ce)brb-g
Tests run at HEAD (35719d52): ``` $ bazel test $(bazel query 'filter(".*getdents.*", //test/syscalls:all)') <snip> //test/syscalls:getdents_test_native PASSED in 0.3s //test/syscalls:getdents_test_runsc_ptrace PASSED in 4.9s //test/syscalls:getdents_test_runsc_ptrace_overlay PASSED in 4.7s //test/syscalls:getdents_test_runsc_ptrace_shared PASSED in 5.2s //test/syscalls:getdents_test_runsc_kvm FAILED in 4.0s ``` Tests run at ab6774ce~1 (6f933a93): ``` $ bazel test $(bazel query 'filter(".*getdents.*", //test/syscalls:all)') //test/syscalls:getdents_test_native PASSED in 0.2s //test/syscalls:getdents_test_runsc_kvm FAILED in 4.2s /usr/local/google/home/brb/.cache/bazel/_bazel_brb/967240a6aae7d353a221d73f4375e038/execroot/__main__/bazel-out/k8-fastbuild/testlogs/test/syscalls/getdents_test_runsc_kvm/test.log //test/syscalls:getdents_test_runsc_ptrace FAILED in 5.3s /usr/local/google/home/brb/.cache/bazel/_bazel_brb/967240a6aae7d353a221d73f4375e038/execroot/__main__/bazel-out/k8-fastbuild/testlogs/test/syscalls/getdents_test_runsc_ptrace/test.log //test/syscalls:getdents_test_runsc_ptrace_overlay FAILED in 4.9s /usr/local/google/home/brb/.cache/bazel/_bazel_brb/967240a6aae7d353a221d73f4375e038/execroot/__main__/bazel-out/k8-fastbuild/testlogs/test/syscalls/getdents_test_runsc_ptrace_overlay/test.log //test/syscalls:getdents_test_runsc_ptrace_shared FAILED in 5.2s /usr/local/google/home/brb/.cache/bazel/_bazel_brb/967240a6aae7d353a221d73f4375e038/execroot/__main__/bazel-out/k8-fastbuild/testlogs/test/syscalls/getdents_test_runsc_ptrace_shared/test.log ``` (I think all runsc_kvm tests are broken on my machine -- I'll rerun them if you can point me at the documentation to set it up)
2019-06-24Return ENOENT when reading /proc/{pid}/task of an exited processchris.zn
There will be a deadloop when we use getdents to read /proc/{pid}/task of an exited process Like this: Process A is running Process B: open /proc/{pid of A}/task Process A exits Process B: getdents /proc/{pid of A}/task Then, process B will fall into deadloop, and return "." and ".." in loops and never ends. This patch returns ENOENT when use getdents to read /proc/{pid}/task if the process is just exited. Signed-off-by: chris.zn <chris.zn@antfin.com>
2019-06-22Implement statx.Nicolas Lacasse
We don't have the plumbing for btime yet, so that field is left off. The returned mask indicates that btime is absent. Fixes #343 PiperOrigin-RevId: 254575752
2019-06-21Fix the logic for sending zero window updates.Bhasker Hariharan
Today we have the logic split in two places between endpoint Read() and the worker goroutine which actually sends a zero window. This change makes it so that when a zero window ACK is sent we set a flag in the endpoint which can be read by the endpoint to decide if it should notify the worker to send a nonZeroWindow update. The worker now does not do the check again but instead sends an ACK and flips the flag right away. Similarly today when SO_RECVBUF is set the SetSockOpt call has logic to decide if a zero window update is required. Rather than do that we move the logic to the worker goroutine and it can check the zeroWindow flag and send an update if required. PiperOrigin-RevId: 254505447
2019-06-21gvisor/fs: getdents returns 0 if offset is equal to FileMaxOffsetAndrei Vagin
FileMaxOffset is a special case when lseek(d, 0, SEEK_END) has been called. PiperOrigin-RevId: 254498777
2019-06-21Remove O(n) lookup on unlink/renameMichael Pratt
Currently, the path tracking in the gofer involves an O(n) lookup of child fidRefs. This causes a significant overhead on unlinks in directories with lots of child fidRefs (<4k). In this transition, pathNode moves from sync.Map to normal synchronized maps. There is a small chance of contention in walk, but the lock is held for a very short time (and sync.Map also had a chance of requiring locking). OTOH, sync.Map makes it very difficult to add a fidRef reverse map. PiperOrigin-RevId: 254489952
2019-06-21Deflake TestSimpleReceive failures due to timeoutsBrad Burlage
This test will occasionally fail waiting to read a packet. From repeated runs, I've seen it up to 1.5s for waitForPackets to complete. PiperOrigin-RevId: 254484627
2019-06-21ext4 block group descriptor implementation in disk layout package.Ayush Ranjan
PiperOrigin-RevId: 254482180
2019-06-21Add //pkg/flipcall.Jamie Liu
Flipcall is a (conceptually) simple local-only RPC mechanism. Compared to unet, Flipcall does not support passing FDs (support for which will be provided out of band by another package), requires users to establish connections manually, and requires user management of concurrency since each connected Endpoint pair supports only a single RPC at a time; however, it improves performance by using shared memory for data (reducing memory copies) and using futexes for control signaling (which is much cheaper than sendto/recvfrom/sendmsg/recvmsg). PiperOrigin-RevId: 254471986
2019-06-21Add list of stuck tasks to panic messageFabricio Voznika
PiperOrigin-RevId: 254450309
2019-06-21Update pathNode documentation to reflect realityMichael Pratt
Neither fidRefs or children are (directly) synchronized by mu. Remove the preconditions that say so. That said, the surrounding does enforce some synchronization guarantees (e.g., fidRef.renameChildTo does not atomically replace the child in the maps). I've tried to note the need for callers to do this synchronization. I've also renamed the maps to what are (IMO) clearer names. As is, it is not obvious that pathNode.fidRefs is a map of *child* fidRefs rather than self fidRefs. PiperOrigin-RevId: 254446965