summaryrefslogtreecommitdiffhomepage
path: root/test/syscalls
AgeCommit message (Collapse)Author
2020-09-09Don't bind loopback to all IPs in an IPv6 subnetGhanan Gowripalan
An earlier change considered the loopback bound to all addresses in an assigned subnet. This should have only be done for IPv4 to maintain compatability with Linux: ``` $ ip addr show dev lo 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group ... link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo valid_lft forever preferred_lft forever inet6 ::1/128 scope host valid_lft forever preferred_lft forever $ ping 2001:db8::1 PING 2001:db8::1(2001:db8::1) 56 data bytes ^C --- 2001:db8::1 ping statistics --- 4 packets transmitted, 0 received, 100% packet loss, time 3062ms $ ping 2001:db8::2 PING 2001:db8::2(2001:db8::2) 56 data bytes ^C --- 2001:db8::2 ping statistics --- 3 packets transmitted, 0 received, 100% packet loss, time 2030ms $ sudo ip addr add 2001:db8::1/64 dev lo $ ping 2001:db8::1 PING 2001:db8::1(2001:db8::1) 56 data bytes 64 bytes from 2001:db8::1: icmp_seq=1 ttl=64 time=0.055 ms 64 bytes from 2001:db8::1: icmp_seq=2 ttl=64 time=0.074 ms 64 bytes from 2001:db8::1: icmp_seq=3 ttl=64 time=0.073 ms 64 bytes from 2001:db8::1: icmp_seq=4 ttl=64 time=0.071 ms ^C --- 2001:db8::1 ping statistics --- 4 packets transmitted, 4 received, 0% packet loss, time 3075ms rtt min/avg/max/mdev = 0.055/0.068/0.074/0.007 ms $ ping 2001:db8::2 PING 2001:db8::2(2001:db8::2) 56 data bytes From 2001:db8::1 icmp_seq=1 Destination unreachable: No route From 2001:db8::1 icmp_seq=2 Destination unreachable: No route From 2001:db8::1 icmp_seq=3 Destination unreachable: No route From 2001:db8::1 icmp_seq=4 Destination unreachable: No route ^C --- 2001:db8::2 ping statistics --- 4 packets transmitted, 0 received, +4 errors, 100% packet loss, time 3070ms ``` Test: integration_test.TestLoopbackAcceptAllInSubnet PiperOrigin-RevId: 329011566
2020-09-09Implement StatFS for various VFS2 filesystems.Rahat Mahmood
This mainly involved enabling kernfs' client filesystems to provide a StatFS implementation. Fixes #3411, #3515. PiperOrigin-RevId: 329009864
2020-09-09Fix EOF handling for splice.Dean Deng
Also, add corresponding EOF tests for splice/sendfile. Discovered by syzkaller. PiperOrigin-RevId: 328975990
2020-09-09fix panic when calling SO_ORIGINAL_DST without initializing iptablesKevin Krakauer
Reported-by: syzbot+074ec22c42305725b79f@syzkaller.appspotmail.com PiperOrigin-RevId: 328963899
2020-09-09Add test demonstrating accept bugTamir Duberstein
Updates #3780. PiperOrigin-RevId: 328922573
2020-09-09Fix vfs2 pipe behavior when splicing to a non-pipe.Dean Deng
Fixes *.sh Java runtime tests, where splice()-ing from a pipe to /dev/zero would not actually empty the pipe. There was no guarantee that the data would actually be consumed on a splice operation unless the output file's implementation of Write/PWrite actually called VFSPipeFD.CopyIn. Now, whatever bytes are "written" are consumed regardless of whether CopyIn is called or not. Furthermore, the number of bytes in the IOSequence for reads is now capped at the amount of data actually available. Before, splicing to /dev/zero would always return the requested splice size without taking the actual available data into account. This change also refactors the case where an input file is spliced into an output pipe so that it follows a similar pattern, which is arguably cleaner anyway. Updates #3576. PiperOrigin-RevId: 328843954
2020-09-09unix: return ECONNREFUSE if a socket file exists but a socket isn't bound to itAndrei Vagin
PiperOrigin-RevId: 328843560
2020-09-09beef up write syscall testsJinmou Li
Added a few tests for write(2) and pwrite(2) 1. Regular Files For write(2) - write zero bytes should not move the offset - write non-zero bytes should increment the offset the exact amount - write non-zero bytes after a lseek() should move the offset the exact amount after the seek - write non-zero bytes with O_APPEND should move the offset the exact amount after original EOF For pwrite(2), offset is not affected when - pwrite zero bytes - pwrite non-zero bytes For EOF, added a test asserting the EOF (indicated by lseek(SEEK_END)) is updated properly after writing non-zero bytes 2. Symlink Added one pwite64() call for symlink that is written as a counterpart of the existing test using pread64()
2020-09-09Fix BadSocketPair for open source.Zach Koopmans
BadSocketPair test will return several errnos (EPREM, ESOCKTNOSUPPORT, EAFNOSUPPORT) meaning the test is just too specific. Checking the syscall fails is appropriate. PiperOrigin-RevId: 328813071
2020-09-09Skip IPv6UDPUnboundSocketNetlinkTest on native linuxGhanan Gowripalan
...while we figure out of we want to consider the loopback interface bound to all IPs in an assigned IPv6 subnet, or not (to maintain compatibility with Linux). PiperOrigin-RevId: 328807974
2020-09-09ip6tables: (de)serialize ip6tables structsKevin Krakauer
More implementation+testing to follow. #3549. PiperOrigin-RevId: 328770160
2020-09-09Fix JobControl tests for open source.Zach Koopmans
ioctl calls with TIOCSCTTY fail if the calling process already has a controlling terminal, which occurs on a 5.4 kernel like our Ubuntu 18 CI. Thus, run tests calling ioctl TTOCSCTTY in clean subprocess. Also, while we're here, switch out non-inclusive master/slave for main/replica. PiperOrigin-RevId: 328756598
2020-09-09tmpfs: Allow xattrs in the trusted namespace if creds has CAP_SYS_ADMIN.Nicolas Lacasse
This is needed to support the overlay opaque attribute. PiperOrigin-RevId: 328552985
2020-09-09Fix SocketPairTest and BadSocketPairTest in opensource.Zach Koopmans
PiperOrigin-RevId: 328467152
2020-09-09Expose basic coverage information to userspace through kcov interface.Dean Deng
In Linux, a kernel configuration is set that compiles the kernel with a custom function that is called at the beginning of every basic block, which updates the memory-mapped coverage information. The Go coverage tool does not allow us to inject arbitrary instructions into basic blocks, but it does provide data that we can convert to a kcov-like format and transfer them to userspace through a memory mapping. Note that this is not a strict implementation of kcov, which is especially tricky to do because we do not have the same coverage tools available in Go that that are available for the actual Linux kernel. In Linux, a kernel configuration is set that compiles the kernel with a custom function that is called at the beginning of every basic block to write program counters to the kcov memory mapping. In Go, however, coverage tools only give us a count of basic blocks as they are executed. Every time we return to userspace, we collect the coverage information and write out PCs for each block that was executed, providing userspace with the illusion that the kcov data is always up to date. For convenience, we also generate a unique synthetic PC for each block instead of using actual PCs. Finally, we do not provide thread-specific coverage data (each kcov instance only contains PCs executed by the thread owning it); instead, we will supply data for any file specified by -- instrumentation_filter. Also, fix issue in nogo that was causing pkg/coverage:coverage_nogo compilation to fail. PiperOrigin-RevId: 328426526
2020-09-09remove iptables sockopt special casesKevin Krakauer
iptables sockopts were kludged into an unnecessary check, this properly relegates them to the {get,set}SockOptIP functions. PiperOrigin-RevId: 328395135
2020-09-09Support SO_LINGER socket option.Nayana Bidari
When SO_LINGER option is enabled, the close will not return until all the queued messages are sent and acknowledged for the socket or linger timeout is reached. If the option is not set, close will return immediately. This option is mainly supported for connection oriented protocols such as TCP. PiperOrigin-RevId: 328350576
2020-09-09Fix TCP_LINGER2 behavior to match linux.Bhasker Hariharan
We still deviate a bit from linux in how long we will actually wait in FIN-WAIT-2. Linux seems to cap it with TIME_WAIT_LEN and it's not completely obvious as to why it's done that way. For now I think we can ignore that and fix it if it really is an issue. PiperOrigin-RevId: 328324922
2020-09-09Consider loopback bound to all addresses in subnetGhanan Gowripalan
When a loopback interface is configurd with an address and associated subnet, the loopback should treat all addresses in that subnet as an address it owns. This is mimicking linux behaviour as seen below: ``` $ ip addr show dev lo 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group ... link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo valid_lft forever preferred_lft forever inet6 ::1/128 scope host valid_lft forever preferred_lft forever $ ping 192.0.2.1 PING 192.0.2.1 (192.0.2.1) 56(84) bytes of data. ^C --- 192.0.2.1 ping statistics --- 2 packets transmitted, 0 received, 100% packet loss, time 1018ms $ ping 192.0.2.2 PING 192.0.2.2 (192.0.2.2) 56(84) bytes of data. ^C --- 192.0.2.2 ping statistics --- 3 packets transmitted, 0 received, 100% packet loss, time 2039ms $ sudo ip addr add 192.0.2.1/24 dev lo $ ip addr show dev lo 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group ... link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo valid_lft forever preferred_lft forever inet 192.0.2.1/24 scope global lo valid_lft forever preferred_lft forever inet6 ::1/128 scope host valid_lft forever preferred_lft forever $ ping 192.0.2.1 PING 192.0.2.1 (192.0.2.1) 56(84) bytes of data. 64 bytes from 192.0.2.1: icmp_seq=1 ttl=64 time=0.131 ms 64 bytes from 192.0.2.1: icmp_seq=2 ttl=64 time=0.046 ms 64 bytes from 192.0.2.1: icmp_seq=3 ttl=64 time=0.048 ms ^C --- 192.0.2.1 ping statistics --- 3 packets transmitted, 3 received, 0% packet loss, time 2042ms rtt min/avg/max/mdev = 0.046/0.075/0.131/0.039 ms $ ping 192.0.2.2 PING 192.0.2.2 (192.0.2.2) 56(84) bytes of data. 64 bytes from 192.0.2.2: icmp_seq=1 ttl=64 time=0.131 ms 64 bytes from 192.0.2.2: icmp_seq=2 ttl=64 time=0.069 ms 64 bytes from 192.0.2.2: icmp_seq=3 ttl=64 time=0.049 ms 64 bytes from 192.0.2.2: icmp_seq=4 ttl=64 time=0.035 ms ^C --- 192.0.2.2 ping statistics --- 4 packets transmitted, 4 received, 0% packet loss, time 3049ms rtt min/avg/max/mdev = 0.035/0.071/0.131/0.036 ms ``` Test: integration_test.TestLoopbackAcceptAllInSubnet PiperOrigin-RevId: 328188546
2020-09-09Add syscall tests for SO_REUSEADDR.Mithun Iyer
Add tests for socket re-bind/listen of client and server sockets with the older connection still in TIME_WAIT state and with SO_REUSEADDR enabled. PiperOrigin-RevId: 327924702
2020-09-09Move udp port exhaustion tests to a 'nogotsan' one.Ting-Yu Wang
It frequently times out under GoTSAN. PiperOrigin-RevId: 327894343
2020-09-09Fix Inotify tests in open source.Zach Koopmans
The order of unlink events (dir event/file event) is undefined, so make tests accept both orderings. PiperOrigin-RevId: 327873316
2020-09-09Skip ElfInterpreterStaticTest for 5.X kernels.Zach Koopmans
gVisor emulates 4.6 kernel versions, and test doesn't work on 5.0 versions (observed on our Ubuntu18.04 image). Skip it. PiperOrigin-RevId: 327845037
2020-09-09Skip socket stress tests on KVM platform.Nicolas Lacasse
They time out. PiperOrigin-RevId: 327830892
2020-09-09Enable strace+debug in syscall testsFabricio Voznika
This is done to ease troubleshooting when tests fail. runsc logs are not stored when tests passe, so this will only affect failing tests and should not increase log storage too badly. PiperOrigin-RevId: 327717551
2020-09-09Skip listening TCP ports when trying to bind a free port.Bhasker Hariharan
PiperOrigin-RevId: 327686558
2020-09-09Fix return for rseq_test.Zach Koopmans
Accept 128 + SIGNAL as well as SIGNAL as valid returns for fork/exec tests. Also, make changes so that test compiles in opensource. Test had compile errors on latest Ubuntu 16.04 image with updated bazel to 3.4.0 (as well as base 2.0) used for Kokoro tests. PiperOrigin-RevId: 327510310
2020-09-09ip6tables: test initial stateKevin Krakauer
Tests that we have the correct initial (empty) state for ip6tables. #3549 PiperOrigin-RevId: 327477657
2020-08-19Skip ECN test for native/linux tests.Zach Koopmans
Skip check for ECN bits in native/linux tests. General advice for the ECN field is to leave the setting to the kernel, so behavior of the test is undefined for different kernels. http://www.masterraghu.com/subjects/np/introduction/unix_network_programming_v1.3/ch07lev1sec6.html PiperOrigin-RevId: 327451414
2020-08-19Return appropriate errors when file locking is unsuccessful.Dean Deng
test_eintr now passes in the Python runtime tests. Updates #3515. PiperOrigin-RevId: 327441081
2020-08-19Fix return value in shm_test for opensourceZach Koopmans
Some machines return 128 + signal for failures. Accept that as a valid result. PiperOrigin-RevId: 327326113
2020-08-19Fix timeval for Socket testZach Koopmans
tv_usec field should be a multiple of 4K to pass in open source on linux/native, so make it one. PiperOrigin-RevId: 327288405
2020-08-19Fix readahead test for opensource.Zach Koopmans
Skip InvalidOffset and InvalidLength for Linux as the test is invalid for later Kernel versions. Add UnsupportedFile test as this check is in all kernel versions. PiperOrigin-RevId: 327248035
2020-08-19Fix return value for MMap Tests in OpenSourceZach Koopmans
Some systems return 128 + errno instead of just errno, which is the case here. PiperOrigin-RevId: 327247836
2020-08-19Fix AllSocketPairTest for open source.Zach Koopmans
Setting timeouts for sockets on GCP images (debian) for usecs only respects multiples of 4K. Set the test with a multiple of 4K with a comment. PiperOrigin-RevId: 327093848
2020-08-19[vfs] Do O_DIRECTORY check after resolving symlinks.Ayush Ranjan
Fixes python runtime test test_glob. Updates #3515 We were checking is the to-be-opened dentry is a dir or not before resolving symlinks. We should check that after resolving symlinks. This was preventing us from opening a symlink which pointed to a directory with O_DIRECTORY. Also added this check in tmpfs and removed a duplicate check. PiperOrigin-RevId: 327085895
2020-08-14Skip UDPMulticast Tests when net interfaces aren't found.Zach Koopmans
PiperOrigin-RevId: 326686761
2020-08-13[vfs2][gofer] Fix file creation flags sent to gofer.Ayush Ranjan
Fixes php runtime test ext/standard/tests/file/readfile_basic.phpt Fixes #3516 fsgofers only want the access mode in the OpenFlags passed to Create(). If more flags are supplied (like O_APPEND in this case), read/write from that fd will fail with EBADF. See runsc/fsgofer/fsgofer.go:WriteAt() VFS2 was providing more than just access modes. So filtering the flags using p9.OpenFlagsModeMask == linux.O_ACCMODE fixes the issue. Gofer in VFS1 also only extracts the access mode flags while making the create RPC. See pkg/sentry/fs/gofer/path.go:Create() Even in VFS2, when we open a handle, we extract out only the access mode flags + O_TRUNC. See third_party/gvisor/pkg/sentry/fsimpl/gofer/handle.go:openHandle() Added a test for this. PiperOrigin-RevId: 326574829
2020-08-13Disable vfs2 for socket_stress_test.Ayush Ranjan
PiperOrigin-RevId: 326553620
2020-08-13Ensure TCP TIME-WAIT is not terminated prematurely.Bhasker Hariharan
Netstack's TIME-WAIT state for a TCP socket could be terminated prematurely if the socket entered TIME-WAIT using shutdown(..., SHUT_RDWR) and then was closed using close(). This fixes that bug and updates the tests to verify that Netstack correctly honors TIME-WAIT under such conditions. Fixes #3106 PiperOrigin-RevId: 326456443
2020-08-12[vfs2][gofer] Return appropriate errors when opening and creating files.Ayush Ranjan
Fixes php test ext/standard/tests/file/touch_variation5.phpt on vfs2. Updates #3516 Also spotted a bug with O_EXCL, where we did not return EEXIST when we tried to open the root of the filesystem with O_EXCL | O_CREAT. Added some more tests for open() corner cases. PiperOrigin-RevId: 326346863
2020-08-10Populate IPPacketInfo with destination addressGhanan Gowripalan
IPPacketInfo.DestinationAddr should hold the destination of the IP packet, not the source. This change fixes that bug. PiperOrigin-RevId: 325910766
2020-08-10Enable VFS2 by default for all syscall tests.Ayush Ranjan
Fixes #2923 PiperOrigin-RevId: 325904734
2020-08-07tcp: change the limit of TCP_LINGER2Andrei Vagin
It was changed in the Linux kernel: commit f0628c524fd188c3f9418e12478dfdfadacba815 Date: Fri Apr 24 16:06:16 2020 +0800 net: Replace the limit of TCP_LINGER2 with TCP_FIN_TIMEOUT_MAX PiperOrigin-RevId: 325493859
2020-08-07Support separate read/write handles in fsimpl/gofer.dentry.Jamie Liu
PiperOrigin-RevId: 325490674
2020-08-07Tolerate EINTR from fallocate() in inotify test.Jamie Liu
PiperOrigin-RevId: 325472312
2020-08-05Add loss recovery option for TCP.Nayana Bidari
/proc/sys/net/ipv4/tcp_recovery is used to enable RACK loss recovery in TCP. PiperOrigin-RevId: 325157807
2020-08-03Add inotify events for fallocate and tests for fallocate/sendfile.Dean Deng
Updates #1479, #2923. PiperOrigin-RevId: 324658826
2020-08-03Merge pull request #3469 from craig08:fuse-fix-test-mount-failedgVisor bot
PiperOrigin-RevId: 324657612
2020-07-31Fix MountFuseFilesystem tests failingCraig Chi
Before kernel version 4.16-rc6, fuse mount is protected by capable(CAP_SYS_ADMIN). After this version, it uses ns_capable(CAP_SYS_ADMIN) to protect. Before the 4.16 kernel, it was not allowed to mount fuse file systems without the global CAP_SYS_ADMIN. Fixes #3360
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-29 11:48:21 +0000