summaryrefslogtreecommitdiffhomepage
path: root/pkg
AgeCommit message (Collapse)Author
2021-02-06Remove linkAddrCacheGhanan Gowripalan
It was replaced by NUD/neighborCache. Fixes #4658. PiperOrigin-RevId: 356085221
2021-02-06Synchronously send packets over pipe link endpointGhanan Gowripalan
Before this change, packets were delivered asynchronously to the remote end of a pipe. This was to avoid a deadlock during link resolution where the stack would attempt to double-lock a mutex (see removed comments in the parent commit for details). As of https://github.com/google/gvisor/commit/4943347137, we do not hold locks while sending link resolution probes so the deadlock will no longer occur. PiperOrigin-RevId: 356066224
2021-02-06Use fine grained locks while sending NDP packetsGhanan Gowripalan
Previously when sending NDP DAD or RS messages, we would hold a shared lock which lead to deadlocks (due to synchronous packet loooping (e.g. pipe and loopback link endpoints)) and lock contention. Writing packets may be an expensive operation which could prevent other goroutines from doing meaningful work if a shared lock is held while writing packets. This change upates the NDP DAD/RS timers to not hold shared locks while sending packets. PiperOrigin-RevId: 356053146
2021-02-06Remove (*stack.Stack).FindNetworkEndpointGhanan Gowripalan
The network endpoints only look for other network endpoints of the same kind. Since the network protocols keeps track of all endpoints, go through the protocol to find an endpoint with an address instead of the stack. PiperOrigin-RevId: 356051498
2021-02-06Use fine grained locks while sending NUD probesGhanan Gowripalan
Previously when sending probe messages, we would hold a shared lock which lead to deadlocks (due to synchronous packet loooping (e.g. pipe and loopback link endpoints)) and lock contention. Writing packets may be an expensive operation which could prevent other goroutines from doing meaningful work if a shared lock is held while writing packets. This change upates the NUD timers to not hold shared locks while sending packets. PiperOrigin-RevId: 356048697
2021-02-06Use embedded mutex pattern in neighbor cache/entryGhanan Gowripalan
Also while I'm here, update neighbor cahce/entry tests to use the stack's RNG instead of creating a neigbor cache/entry specific one. PiperOrigin-RevId: 356040581
2021-02-06Unexpose NICGhanan Gowripalan
The NIC structure is not to be used outside of the stack package directly. PiperOrigin-RevId: 356036737
2021-02-06Check local address directly through NICGhanan Gowripalan
Network endpoints that wish to check addresses on another NIC-local network endpoint may now do so through the NetworkInterface. This fixes a lock ordering issue between NIC removal and link resolution. Before this change: NIC Removal takes the stack lock, neighbor cache lock then neighbor entries' locks. When performing IPv4 link resolution, we take the entry lock then ARP would try check IPv4 local addresses through the stack which tries to obtain the stack's lock. Now that ARP can check IPv4 addreses through the NIC, we avoid the lock ordering issue, while also removing the need for stack to lookup the NIC. PiperOrigin-RevId: 356034245
2021-02-05Batch write packets after iptables checksGhanan Gowripalan
After IPTables checks a batch of packets, we can write packets that are not dropped or locally destined as a batch instead of individually. This previously caused a bug since WritePacket* functions expect to take ownership of passed PacketBuffer{List}. WritePackets assumed the list of PacketBuffers will not be invalidated when calling WritePacket for each PacketBuffer in the list, but this is not true. WritePacket may add the passed PacketBuffer into a different list which would modify the PacketBuffer in such a way that it no longer points to the next PacketBuffer to write. Example: Given a PB list of PB_a -> PB_b -> PB_c WritePackets may be iterating over the list and calling WritePacket for each PB. When WritePacket takes PB_a, it may add it to a new list which would update pointers such that PB_a no longer points to PB_b. Test: integration_test.TestIPTableWritePackets PiperOrigin-RevId: 355969560
2021-02-05Replace TaskFromContext(ctx).Kernel() with KernelFromContext(ctx)Ting-Yu Wang
Panic seen at some code path like control.ExecAsync where ctx does not have a Task. Reported-by: syzbot+55ce727161cf94a7b7d6@syzkaller.appspotmail.com PiperOrigin-RevId: 355960596
2021-02-05[vfs] Handle `.` and `..` as last path component names in kernfs Rename.Ayush Ranjan
According to vfs.FilesystemImpl.RenameAt documentation: - If the last path component in rp is "." or "..", and opts.Flags contains RENAME_NOREPLACE, RenameAt returns EEXIST. - If the last path component in rp is "." or "..", and opts.Flags does not contain RENAME_NOREPLACE, RenameAt returns EBUSY. Reported-by: syzbot+6189786e64fe13fe43f8@syzkaller.appspotmail.com PiperOrigin-RevId: 355959266
2021-02-05Refactor locally delivered packetsGhanan Gowripalan
Make it clear that failing to parse a looped back is not a packet sending error but a malformed received packet error. FindNetworkEndpoint returns nil when no network endpoint is found instead of an error. PiperOrigin-RevId: 355954946
2021-02-04Lock ConnTrack before initializing bucketsGhanan Gowripalan
PiperOrigin-RevId: 355751801
2021-02-04Move getcpu() to core filter listMichael Pratt
Some versions of the Go runtime call getcpu(), so add it for compatibility. The hostcpu package already uses getcpu() on arm64. PiperOrigin-RevId: 355717757
2021-02-04Call kernfs.syntheticDir.InitRefs() on creation.Nicolas Lacasse
PiperOrigin-RevId: 355675900
2021-02-03[vfs] Make sticky bit check consistent with Linux.Ayush Ranjan
Our implementation of vfs.CheckDeleteSticky was not consistent with Linux, specifically not consistent with fs/linux.h:check_sticky(). One of the biggest differences was that the vfs implementation did not allow the owner of the sticky directory to delete files inside it that belonged to other users. This change makes our implementation consistent with Linux. Also adds an integration test to check for this. This bug is also present in VFS1. Updates #3027 PiperOrigin-RevId: 355557425
2021-02-03Add a function to enable RACK in tests.Nayana Bidari
- Adds a function to enable RACK in tests. - RACK update functions are guarded behind the flag tcpRecovery. PiperOrigin-RevId: 355435973
2021-02-02Remove call to Notify from pipe.VFSPipeFD.CopyOutFrom.Jamie Liu
This was missed in cl/351911375; pipe.VFSPipeFD.SpliceFromNonPipe already calls Notify. PiperOrigin-RevId: 355246655
2021-02-02Rename HandleNDupAcks in TCP.Nayana Bidari
Rename HandleNDupAcks() to HandleLossDetected() as it will enter this when is detected after: - reorder window expires and TLP (in case of RACK) - dupAckCount >= 3 PiperOrigin-RevId: 355237858
2021-02-02Stub out basic `runsc events --stat` CPU functionalityKevin Krakauer
Because we lack gVisor-internal cgroups, we take the CPU usage of the entire pod and divide it proportionally according to sentry-internal usage stats. This fixes `kubectl top pods`, which gets a pod's CPU usage by summing the usage of its containers. Addresses #172. PiperOrigin-RevId: 355229833
2021-02-02Move ring0 package.Adin Scannell
This allows the package to serve as a general purpose ring0 support package, as opposed to being bound to specific sentry platforms. Updates #5039 PiperOrigin-RevId: 355220044
2021-02-02Add ETIMEDOUT to partial result listFabricio Voznika
Reported-by: syzbot+db8d83f93b84fcb84374@syzkaller.appspotmail.com PiperOrigin-RevId: 355213994
2021-02-02Add support for rate limiting out of window ACKs.Bhasker Hariharan
Netstack today will send dupACK's with no rate limit for incoming out of window segments. This can result in ACK loops for example if a TCP socket connects to itself (actually permitted by TCP). Where the ACK sent in response to packets being out of order itself gets considered as an out of window segment resulting in another ACK being generated. PiperOrigin-RevId: 355206877
2021-02-02Minor page tables improvements.Adin Scannell
* Make split safe. * Enable looking up next valid address. * Support mappings with !accessType.Any(), distinct from unmap. These changes allow for the use of pagetables in low-level OS packages, such as ring0, and allow for the use of pagetables for more generic address space reservation (by writing entries with no access specified). Updates #5039 PiperOrigin-RevId: 355109016
2021-02-01Refactor HandleControlPacket/SockErrorGhanan Gowripalan
...to remove the need for the transport layer to deduce the type of error it received. Rename HandleControlPacket to HandleError as HandleControlPacket only handles errors. tcpip.SockError now holds a tcpip.SockErrorCause interface that different errors can implement. PiperOrigin-RevId: 354994306
2021-01-31Default to NUD/neighborCache instead of linkAddrCacheGhanan Gowripalan
This change flips gvisor to use Neighbor unreachability detection by default to populate the neighbor table as defined by RFC 4861 section 7. Although RFC 4861 is targeted at IPv6, the same algorithm is used for link resolution on IPv4 networks using ARP. Integrators may still use the legacy link address cache by setting stack.Options.UseLinkAddrCache to true; stack.Options.UseNeighborCache is now unused and will be removed. A later change will remove linkAddrCache and associated code. Updates #4658. PiperOrigin-RevId: 354850531
2021-01-31Use closure for IPv6 testContext cleanupGhanan Gowripalan
PiperOrigin-RevId: 354827491
2021-01-31Remove NICs before closing their link endpointsGhanan Gowripalan
...in IPv6 ICMP tests. A channel link endpoint's channel is closed when the link endpoint is closed. When the stack tries to send packets through a NIC with a closed channel endpoint, a panic will occur when attempting to write to a closed channel (https://golang.org/ref/spec#Close). To make sure the stack does not try to send packets through a NIC, we remove it. PiperOrigin-RevId: 354822085
2021-01-31Use different neighbor tables per network endpointGhanan Gowripalan
This stores each protocol's neighbor state separately. This change also removes the need for each neighbor entry to keep track of their own link address resolver now that all the entries in a cache will use the same resolver. PiperOrigin-RevId: 354818155
2021-01-31Hide neighbor table kind from NetworkEndpointGhanan Gowripalan
The network endpoint should not need to have logic to handle different kinds of neighbor tables. Network endpoints can let the NIC know about differnt neighbor discovery messages and let the NIC decide which table to update. This allows us to remove the LinkAddressCache interface. PiperOrigin-RevId: 354812584
2021-01-30Extract route table from Stack lockTamir Duberstein
PiperOrigin-RevId: 354746864
2021-01-30Implement LinkAddressResolver on NetworkEndpointsGhanan Gowripalan
This removes the need to provide the link address request with the NIC the request is being performed on since the NetworkEndpoints already have a reference to the NIC. PiperOrigin-RevId: 354721940
2021-01-29Make fragmentation return a reassembled PacketBufferTing-Yu Wang
This allows later decoupling of the backing network buffer implementation. PiperOrigin-RevId: 354643297
2021-01-29Clear IGMPv1 present flag on NIC downGhanan Gowripalan
This is dynamic state that can be re-learned when the NIC comes back up. Test: ipv4_test.TestIgmpV1Present PiperOrigin-RevId: 354630921
2021-01-29Refresh delayed report timers on query messagesGhanan Gowripalan
...as per As per RFC 2236 section 3 page 3 (for IGMPv2) and RFC 2710 section 4 page 5 (for MLDv1). See comments in code for more details. Test: ip_test.TestHandleQuery PiperOrigin-RevId: 354603068
2021-01-29- Add more comments for the TCP_INFO struct fields.Nayana Bidari
PiperOrigin-RevId: 354595623
2021-01-29Fix deadlock in specialFileFD.pwriteFabricio Voznika
When file is regular and metadata cache is authoritative, metadata lock is taken. The code deadlocks trying to acquire the metadata lock again to update time stampts. PiperOrigin-RevId: 354584594
2021-01-29Discard invalid Neighbor AdvertisementsPeter Johnston
...per RFC 4861 s7.1.2. Startblock: has LGTM from sbalana and then add reviewer ghanan PiperOrigin-RevId: 354539026
2021-01-28Avoid locking when route doesn't require resolutionGhanan Gowripalan
When a route does not need to resolve a remote link address to send a packet, avoid having to obtain the pending packets queue's lock. PiperOrigin-RevId: 354456280
2021-01-28RACK: Update reorder window.Nayana Bidari
After receiving an ACK(cumulative or selective), RACK will update the reorder window which is used as a settling time before marking the packet as lost. This change will add an init function to initialize the variables in RACK and also store the reference to sender in rackControl. The reorder window is calculated as per rfc: https://tools.ietf.org/html/draft-ietf-tcpm-rack-08#section-7.2 Step 4. PiperOrigin-RevId: 354453528
2021-01-28Acquire entry lock with cache lock heldTamir Duberstein
Avoid a race condition in which an entry is acquired while it is being evicted by overlapping the entry lock with the cache lock. PiperOrigin-RevId: 354452639
2021-01-28Change tcpip.Error to an interfaceTamir Duberstein
This makes it possible to add data to types that implement tcpip.Error. ErrBadLinkEndpoint is removed as it is unused. PiperOrigin-RevId: 354437314
2021-01-28Propagate reader error in ReadFromTamir Duberstein
This was removed in 6c0e1d9cfe6adbfbb32e7020d6426608ac63ad37 but turns out to be crucial to prevent flaky behaviour in sendfile. PiperOrigin-RevId: 354434144
2021-01-28Do not use clockwork for faketimeGhanan Gowripalan
Clockwork does not support timers being reset/stopped from different goroutines. Our current use of clockwork causes data races and gotsan complains about clockwork. This change uses our own implementation of faketime, avoiding data races. PiperOrigin-RevId: 354428208
2021-01-28[vfs] Fix rename implementation in OrderedChildren.Ayush Ranjan
Fixes #3027 as there is just 1 writable user using OrderedChildren's rename, unlink and rmdir (kernfs.syntheticDirectory) but it doesn't support the sticky bit yet. Fuse which is the other writable user implements its own Inode operations. PiperOrigin-RevId: 354386522
2021-01-28Add O_PATH support in vfs2gVisor bot
PiperOrigin-RevId: 354367665
2021-01-28Internal change.Michael Pratt
PiperOrigin-RevId: 354327882
2021-01-28Respect SO_BINDTODEVICE in unconnected UDP writesMarina Ciocea
Previously, sending on an unconnected UDP socket would ignore the SO_BINDTODEVICE option. Send on the configured interface when an UDP socket is bound to an interface through setsockop SO_BINDTODEVICE. Add packetimpact tests exercising UDP reads and writes with every combination of bound/unbound, broadcast/multicast/unicast destination, and bound/not-bound to device. PiperOrigin-RevId: 354299670
2021-01-27Confirm neighbor reachability with TCP ACKsGhanan Gowripalan
As per RFC 4861 section 7.3.1, A neighbor is considered reachable if the node has recently received a confirmation that packets sent recently to the neighbor were received by its IP layer. Positive confirmation can be gathered in two ways: hints from upper-layer protocols that indicate a connection is making "forward progress", or receipt of a Neighbor Advertisement message that is a response to a Neighbor Solicitation message. This change adds support for TCP to let the IP/link layers know that a neighbor is reachable. Test: integration_test.TestTCPConfirmNeighborReachability PiperOrigin-RevId: 354222833
2021-01-27Rename anonymous struct "mu"Tamir Duberstein
This clarifies that there is a lock involved. PiperOrigin-RevId: 354213848
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-19 22:23:30 +0000