summaryrefslogtreecommitdiffhomepage
path: root/runsc/fsgofer
AgeCommit message (Collapse)Author
2018-09-05Add MADVISE to fsgofer seccomp profileFabricio Voznika
PiperOrigin-RevId: 211686037 Change-Id: I0e776ca760b65ba100e495f471b6e811dbd6590a
2018-08-30Add seccomp filter to fsgoferFabricio Voznika
PiperOrigin-RevId: 211011542 Change-Id: Ib5a83a00f8eb6401603c6fb5b59afc93bac52558
2018-08-27runsc: fsgofer should return a unique QID.Path for each file.Nicolas Lacasse
Previously, we were only using the host inode id as the QID path. But the host filesystem can have multiple devices with conflicting inode ids. This resulted in duplicate inode ids in the sentry. This CL generates a unique QID for each <host inode, host device> pair. PiperOrigin-RevId: 210424813 Change-Id: I16d106f61c7c8f910c0da4ceec562a010ffca2fb
2018-08-27Put fsgofer inside chrootFabricio Voznika
Now each container gets its own dedicated gofer that is chroot'd to the rootfs path. This is done to add an extra layer of security in case the gofer gets compromised. PiperOrigin-RevId: 210396476 Change-Id: Iba21360a59dfe90875d61000db103f8609157ca0
2018-08-24Add option to panic gofer if writes are attempted over RO mountsFabricio Voznika
This is used when '--overlay=true' to guarantee writes are not sent to gofer. PiperOrigin-RevId: 210116288 Change-Id: I7616008c4c0e8d3668e07a205207f46e2144bf30
2018-08-22runsc: De-flakes container_test TestMultiContainerSanity.Kevin Krakauer
The bug was caused by os.File's finalizer, which closes the file. Because fsgofer.serve() was passed a file descriptor as an int rather than a os.File, callers would pass os.File.Fd(), and the os.File would go out of scope. Thus, the file would get GC'd and finalized nondeterministically, causing failures when the file was used. PiperOrigin-RevId: 209861834 Change-Id: Idf24d5c1f04c9b28659e62c97202ab3b4d72e994
2018-08-15runsc fsgofer: Support dynamic serving of filesystems.Kevin Krakauer
When multiple containers run inside a sentry, each container has its own root filesystem and set of mounts. Containers are also added after sentry boot rather than all configured and known at boot time. The fsgofer needs to be able to serve the root filesystem of each container. Thus, it must be possible to add filesystems after the fsgofer has already started. This change: * Creates a URPC endpoint within the gofer process that listens for requests to serve new content. * Enables the sentry, when starting a new container, to add the new container's filesystem. * Mounts those new filesystems at separate roots within the sentry. PiperOrigin-RevId: 208903248 Change-Id: Ifa91ec9c8caf5f2f0a9eead83c4a57090ce92068
2018-07-30Cleans up files created if there is a failure.Brielle Broder
PiperOrigin-RevId: 206674267 Change-Id: Ifc4eb19e0882e8bed566e9c553af910925fe6ae2
2018-06-04Make fsgofer attach more strictFabricio Voznika
Refuse to mount paths with "." and ".." in the path to prevent a compromised Sentry to mount "../../secrets". Only allow Attach to be called once per mount point. PiperOrigin-RevId: 199225929 Change-Id: I2a3eb7ea0b23f22eb8dde2e383e32563ec003bd5
2018-05-10Fix failure to rename directoryFabricio Voznika
os.Rename validates that the target doesn't exist, which is different from syscall.Rename which replace the target if both are directories. fsgofer needs the syscall behavior. PiperOrigin-RevId: 196194630 Change-Id: I87d08cad88b5ef310b245cd91647c4f5194159d8
2018-05-10Open file as read-write when mount points to a fileFabricio Voznika
This is to allow files mapped directly, like /etc/hosts, to be writable. Closes #40 PiperOrigin-RevId: 196155920 Change-Id: Id2027e421cef5f94a0951c3e18b398a77c285bbd
2018-05-04Run gofmt -s on everythingIan Gudger
PiperOrigin-RevId: 195469901 Change-Id: I66d5c7a334bbb8b47e40d266a2661291c2d91c7f
2018-04-28Check in gVisor.Googler
PiperOrigin-RevId: 194583126 Change-Id: Ica1d8821a90f74e7e745962d71801c598c652463
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-06 17:42:59 +0000