summaryrefslogtreecommitdiffhomepage
AgeCommit message (Collapse)Author
2019-12-05Reduce flakiness under gotsan runs.Bhasker Hariharan
TcpPortReuseMultiThread creates lots of connections which result in a lot of goroutines in the sentry. This can cause gotsan runs to take really long and timeout. Increasing listen backlog and reducing number of connections should help the connections complete faster as well as reduce the number of goroutines that gotsan needs to track. PiperOrigin-RevId: 284046018
2019-12-05Create correct file for /proc/[pid]/task/[tid]/ioZach Koopmans
PiperOrigin-RevId: 284038840
2019-12-05Add vfs.CheckSetStat() and its dependencies.Jamie Liu
PiperOrigin-RevId: 284033820
2019-12-05Add a type to represent the NDP Recursive DNS Server optionGhanan Gowripalan
This change adds a type to represent the NDP Recursive DNS Server option, as defined by RFC 8106 section 5.1. PiperOrigin-RevId: 284005493
2019-12-05Explicitly export files needed by other packagesgVisor bot
PiperOrigin-RevId: 283955946
2019-12-04Allow syscall tests to run with hostinet.Dean Deng
Fixes #1207 PiperOrigin-RevId: 283914438
2019-12-04Remove TODO since we don't plan to support debug registersFabricio Voznika
PiperOrigin-RevId: 283828423
2019-12-03Clean up readv_socket test suite.Dean Deng
Get rid of the SocketTest class, which is only extended by ReadvSocketTest. Also, get rid of TCP sockets (which were unused anyway) from readv_socket.cc. This is a very old test suite that isn't the right place for TCP loopback tests. PiperOrigin-RevId: 283672772
2019-12-03Point TODO to gvisor.devFabricio Voznika
PiperOrigin-RevId: 283657725
2019-12-03net/udp: return a local route address as the bound-to addressAndrei Vagin
If the socket is bound to ANY and connected to a loopback address, getsockname() has to return the loopback address. Without this fix, getsockname() returns ANY. PiperOrigin-RevId: 283647781
2019-12-03Fix printing /proc/[pid]/io for /proc/[pid]/task/[tid]/io.Zach Koopmans
PiperOrigin-RevId: 283630669
2019-12-03Fix panic due to early transition to Closed.Bhasker Hariharan
The code in rcv.consumeSegment incorrectly transitions to CLOSED state from LAST-ACK before the final ACK for the FIN. Further if receiving a segment changes a socket to a closed state then we should not invoke the sender as the socket is now closed and sending any segments is incorrect. PiperOrigin-RevId: 283625300
2019-12-03strace: don't create a slice with a negative valueAndrei Vagin
PiperOrigin-RevId: 283613824
2019-12-03Remove watchdog TODOFabricio Voznika
I have not seen a false positive stuck task yet. Biggest offender was whitelistfs which is going away. PiperOrigin-RevId: 283613064
2019-12-03Point TODOs to gvisor.devFabricio Voznika
PiperOrigin-RevId: 283610781
2019-12-03Add RunfilesPath to test_utilMichael Pratt
A few tests have their own ad-hoc implementations. Add a single common one. PiperOrigin-RevId: 283601666
2019-12-03Remove TODO for obsolete bug.Zach Koopmans
PiperOrigin-RevId: 283571456
2019-12-03Support IP_TOS and IPV6_TCLASS socket options for hostinet sockets.Dean Deng
There are two potential ways of sending a TOS byte with outgoing packets: including a control message in sendmsg, or setting the IP_TOS/IPV6_TCLASS socket options (for IPV4 and IPV6 respectively). This change lets hostinet support the latter. Fixes #1188 PiperOrigin-RevId: 283550925
2019-12-02Import benchmark-tools to main repository.Zach Koopmans
This has adapted for use with bazel from the original commit a26e93769ebefd82593a43e22fb13a09717cfa6d. In particular, the style has been made consistent with internal python style guidelines, and the packages (including the main entrypoint) have been refactored in order to allow bazel testing targets. PiperOrigin-RevId: 283484433
2019-12-02Allow non-unique UIDs in bazel docker containersIan Lewis
Allow non-unique UIDs in the bazel docker container in order to avoid failures using host UIDs that are already present in the image. Issue #1267 PiperOrigin-RevId: 283456369
2019-12-02test/syscal: Don't skip ClockGettime.CputimeIdAndrei Vagin
We skipped it due to the issue in the golang scheduler which has been fixed in go1.13. PiperOrigin-RevId: 283432226
2019-12-02Support sending IP_TOS and IPV6_TCLASS control messages with hostinet sockets.Dean Deng
There are two potential ways of sending a TOS byte with outgoing packets: including a control message in sendmsg, or setting the IP_TOS/IPV6_TCLASS socket options (for IPV4 and IPV6 respectively). This change lets hostinet support the former. PiperOrigin-RevId: 283346737
2019-12-02Fix typo, s/Convertable/Convertible/gJay Zhuang
PiperOrigin-RevId: 283345791
2019-12-02Port tests in udp_socket.cc to FuchsiaJay Zhuang
Separate out a test in udp_socket.cc that depends on <linux/errqueue.h> so the rest of the tests can run on Fuchsia. PiperOrigin-RevId: 283322633
2019-11-28Test handling segments on completed but not yet accepted TCP connectionsGhanan Gowripalan
This change does not introduce any new features, or modify existing ones. This change tests handling TCP segments right away for connections that were completed from a listening endpoint. PiperOrigin-RevId: 282986457
2019-11-27Add support for receiving TOS and TCLASS control messages in hostinet.Dean Deng
This involves allowing getsockopt/setsockopt for the corresponding socket options, as well as allowing hostinet to process control messages received from the actual recvmsg syscall. PiperOrigin-RevId: 282851425
2019-11-27Add floating point exception testsMichael Pratt
PiperOrigin-RevId: 282828273
2019-11-26Allow open(O_TRUNC) and (f)truncate for proc files.Ian Lewis
This allows writable proc and devices files to be opened with O_CREAT|O_TRUNC. This is encountered most frequently when interacting with proc or devices files via the command line. e.g. $ echo 8192 1048576 4194304 > /proc/sys/net/ipv4/tcp_rmem Also adds a test to test the behavior of open(O_TRUNC), truncate, and ftruncate on named pipes. Fixes #1116 PiperOrigin-RevId: 282677425
2019-11-26Merge pull request #981 from tanjianfeng:fix-898gVisor bot
PiperOrigin-RevId: 282669859
2019-11-26Mark execveat as supported for linux64_arm64.Dean Deng
PiperOrigin-RevId: 282667122
2019-11-25Minor VFS2 interface changes.Jamie Liu
- Remove the Filesystem argument from DentryImpl.*Ref(); in general DentryImpls that need the Filesystem for reference counting will probably also need it for other interface methods that don't plumb Filesystem, so it's easier to just store a pointer to the filesystem in the DentryImpl. - Add a pointer to the VirtualFilesystem to Filesystem, which is needed by the gofer client to disown dentries for cache eviction triggered by dentry reference count changes. - Rename FilesystemType.NewFilesystem to GetFilesystem; in some cases (e.g. sysfs, cgroupfs) it's much cleaner for there to be only one Filesystem that is used by all mounts, and in at least one case (devtmpfs) it's visibly incorrect not to do so, so NewFilesystem doesn't always actually create and return a *new* Filesystem. - Require callers of FileDescription.Init() to increment Mount/Dentry references. This is because the gofer client may, in the OpenAt() path, take a reference on a dentry with 0 references, which is safe due to synchronization that is outside the scope of this CL, and it would be safer to still have its implementation of DentryImpl.IncRef() check for an increment for 0 references in other cases. - Add FileDescription.TryIncRef. This is used by the gofer client to take references on "special file descriptions" (FDs for files such as pipes, sockets, and devices), which use per-FD handles (fids) instead of dentry-shared handles, for sync() and syncfs(). PiperOrigin-RevId: 282473364
2019-11-25Implement basic umounting for vfs2.Jamie Liu
This is required to test filesystems with a non-trivial implementation of FilesystemImpl.Release(). Propagation isn't handled yet, and umount isn't yet plumbed out to VirtualFilesystem.UmountAt(), but otherwise the implementation of umount is believed to be correct. - Move entering mountTable.seq writer critical sections to callers of mountTable.{insert,remove}Seqed. This is required since umount(2) must ensure that no new references are taken on the candidate mount after checking that it isn't busy, which is only possible by entering a vfs.mountTable.seq writer critical section before the check and remaining in it until after VFS.umountRecursiveLocked() is complete. (Linux does the same thing: fs/namespace.c:do_umount() => lock_mount_hash(), fs/pnode.c:propagate_mount_busy(), umount_tree(), unlock_mount_hash().) - It's not possible for dentry deletion to umount while only holding VFS.mountMu for reading, but it's also very unappealing to hold VFS.mountMu exclusively around e.g. gofer unlink RPCs. Introduce dentry.mu to avoid these problems. This means that VFS.mountMu is never acquired for reading, so change it to a sync.Mutex. PiperOrigin-RevId: 282444343
2019-11-25Fix typo in go_branch.sh script.Adin Scannell
With the ticks, the command `master` is actually be run and the output included (which is nothing). This is confusing, as we actually mean to say "master" in the description of the Go branch. PiperOrigin-RevId: 282426081
2019-11-25Use mount hints to determine FileAccessTypeFabricio Voznika
PiperOrigin-RevId: 282401165
2019-11-25Internal change.Adin Scannell
PiperOrigin-RevId: 282396322
2019-11-25Merge pull request #1176 from xiaobo55x:runsc_bootgVisor bot
PiperOrigin-RevId: 282382564
2019-11-25Set transport and network headers on outbound packets.Kevin Krakauer
These are necessary for iptables to read and parse headers for packet filtering. PiperOrigin-RevId: 282372811
2019-11-25Fix panic in sniffer.Kevin Krakauer
Packets written via SOCK_RAW are guaranteed to have network headers, but not transport headers. Check first whether there are enough bytes left in the packet to contain a transport header before attempting to parse it. PiperOrigin-RevId: 282363895
2019-11-23Cleanup visibility.Adin Scannell
PiperOrigin-RevId: 282194656
2019-11-23gofer: reduce CPU usage on GC as of frequent readdirJianfeng Tan
Refer to golang mallocgc(), each time of allocating an object > 32 KB, a gc will be triggered. When we do readdir, sentry always passes 65535, which leads to a malloc of 65535 * sizeof(p9.Direnta) > 32 KB. Considering we already use slice append, let's avoid defining the capability for this slide. Command for test: Before this change: (container)$ time tree linux-5.3.1 > /dev/null real 0m54.272s user 0m2.010s sys 0m1.740s (CPU usage of Gofer: ~30 cores) (host)$ perf top -p <pid-of-gofer> 42.57% runsc [.] runtime.gcDrain 23.41% runsc [.] runtime.(*lfstack).pop 9.74% runsc [.] runtime.greyobject 8.06% runsc [.] runtime.(*lfstack).push 4.33% runsc [.] runtime.scanobject 1.69% runsc [.] runtime.findObject 1.12% runsc [.] runtime.findrunnable 0.69% runsc [.] runtime.runqgrab ... (host)$ mkdir test && cd test (host)$ for i in `seq 1 65536`; do mkdir $i; done (container)$ time ls test/ > /dev/null real 2m10.934s user 0m0.280s sys 0m4.260s (CPU usage of Gofer: ~1 core) After this change: (container)$ time tree linux-5.3.1 > /dev/null real 0m22.465s user 0m1.270s sys 0m1.310s (CPU usage of Gofer: ~1 core) $ perf top -p <pid-of-gofer> 20.57% runsc [.] runtime.gcDrain 7.15% runsc [.] runtime.(*lfstack).pop 4.11% runsc [.] runtime.scanobject 3.78% runsc [.] runtime.greyobject 2.78% runsc [.] runtime.(*lfstack).push ... (host)$ mkdir test && cd test (host)$ for i in `seq 1 65536`; do mkdir $i; done (container)$ time ls test/ > /dev/null real 0m13.338s user 0m0.190s sys 0m3.980s (CPU usage of Gofer: ~0.8 core) Fixes #898 Signed-off-by: Jianfeng Tan <henry.tjf@antfin.com>
2019-11-22Internal change.Adin Scannell
PiperOrigin-RevId: 282068093
2019-11-22Force timezone initialization before filter installationMichael Pratt
The first use of time.Local (usually via time.Time.Date, et. al) performs initialization of the local timezone, which involves open several tzdata files from the host. Since filter installation disallows open, we should explicitly force this initialization rather than implicitly depending on the first logging (or other time) call occurring before filter installation. PiperOrigin-RevId: 282053121
2019-11-22Store SO_BINDTODEVICE state at bind.Ian Gudger
This allows us to ensure that the correct port reservation is released. Fixes #1217 PiperOrigin-RevId: 282048155
2019-11-22Use PacketBuffers with GSO.Kevin Krakauer
PiperOrigin-RevId: 282045221
2019-11-22Add segment dequeue check while emptying segment queue.Mithun Iyer
PiperOrigin-RevId: 282023891
2019-11-22enable ring0/pagetables to support arm64lubinszARM
Signed-off-by: Bin Lu <bin.lu@arm.com> COPYBARA_INTEGRATE_REVIEW=https://github.com/google/gvisor/pull/891 from lubinszARM:pr_pagetable 2385de75a8662af3ab1ae289dd74dd0e5dcfaf66 PiperOrigin-RevId: 282013224
2019-11-22tests: include sys/socket.h before linux/if_arp.hAndrei Vagin
This is how it has to be accoding to the man page. PiperOrigin-RevId: 281998068
2019-11-21Import and structure cleanup.Adin Scannell
PiperOrigin-RevId: 281795269
2019-11-20Use a GitHub credential for tagging a release.Adin Scannell
PiperOrigin-RevId: 281617882
2019-11-20Pass OpenTruncate to gofer in Open call when opening file with O_TRUNC.Nicolas Lacasse
Note that the Sentry still calls Truncate() on the file before calling Open. A new p9 version check was added to ensure that the p9 server can handle the the OpenTruncate flag. If not, then the flag is stripped before sending. PiperOrigin-RevId: 281609112
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-26 13:24:59 +0000