Age | Commit message (Collapse) | Author |
|
PiperOrigin-RevId: 393095246
|
|
.. by advancing the clock so that NowMonotonic does not return 0.
PiperOrigin-RevId: 393005373
|
|
PiperOrigin-RevId: 393004533
|
|
PiperOrigin-RevId: 392982220
|
|
Add logs to big query for benchmarks. This is to debug the actual
sent data from BQ so it can be compared to the data produced by
the benchmark run.
Current logs end with "Sending benchmarks", which is ambiguous.
Log success to make sure they are sent.
PiperOrigin-RevId: 392944104
|
|
PiperOrigin-RevId: 392774712
|
|
Some tcp unit tests are affected by this change:
- Some retransmission tests assumed RTO=1s when connection is established. This
is no longer true because minRTO was set to 3s in tests so now RTO becomes 3s
after the first updateRTO call. Set minRTO=1s for these tests.
- Some RACK enabled tests are affected because now that RTT is initialized, and
the estimated RTT is quite small, spurious TLP might be sent out and causing
flakes, introduce an artificial delay for these tests so that the estimated
RTT is larger.
PiperOrigin-RevId: 392768725
|
|
PiperOrigin-RevId: 392760305
|
|
PiperOrigin-RevId: 392554743
|
|
PiperOrigin-RevId: 392523879
|
|
There is no reason to limit this to debian paths.
PiperOrigin-RevId: 392469415
|
|
Right now, the first slot starts with an address of a memory region and its size is faultBlockSize,
but the second slot starts with (physicalStart + faultBlockSize) & faultBlockMask.
It means they will overlap if a start address of a memory region are not aligned to faultBlockSize.
The kernel doesn't allow to add overlapped regions, but we ignore the EEXIST error.
Signed-off-by: Andrei Vagin <avagin@google.com>
|
|
PiperOrigin-RevId: 392102898
|
|
The old implementation was mostly correct but error prone - making way for the
issue in question here. In its error path, it would leak the intermediate file
being walked. Each return/break needed explicit cleanup.
This change implements a more clean way to cleaning up intermediate directories.
If the code were to evolve to be more complex, it would still work.
PiperOrigin-RevId: 392102826
|
|
We cannot hold mm.aioManager.mu while calling MUnmap, because MUnmap attempts
to aquire mm.mappingMu. This violates the lock order as documented in mm/mm.go.
PiperOrigin-RevId: 392102472
|
|
PiperOrigin-RevId: 392078690
|
|
PiperOrigin-RevId: 392036519
|
|
`tools/go_mod.sh` is not in the repo. In order to update the
WORKSPACE dependencies, we can use the same gazelle command in BUILD file.
Also changed `go mod get` to `go get`, the former does not exist
anymore.
Signed-off-by: Daniel Dao <dqminh89@gmail.com>
|
|
Add an LRU cache to cache verity dentries when ref count drop to 0. This
way we don't need to hash and verify the previous opened files or
directories each time.
PiperOrigin-RevId: 391880157
|
|
Read all data into memory in one Read call and verify them block by
block instead of read each block during verification. This is for
performance purpose to avoid invoking multiple syscalls.
PiperOrigin-RevId: 391877937
|
|
The rationale given for using buffered copies is still valid, but it's unclear
whether holding MM locks or allocating buffers is better in practice, and the
former is at least consistent with gofer.regularFileFD (and VFS1), making
performance easier to reason about.
PiperOrigin-RevId: 391877913
|
|
...to match Linux behaviour.
We can see evidence of Linux representing loopback as an ethernet-based
device below:
```
# EUI-48 based MAC addresses.
$ ip link show lo
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
# tcpdump showing ethernet frames when sniffing loopback and logging the
# link-type as EN10MB (Ethernet).
$ sudo tcpdump -i lo -e -c 2 -n
tcpdump: verbose output suppressed, use -v[v]... for full protocol decode
listening on lo, link-type EN10MB (Ethernet), snapshot length 262144 bytes
03:09:05.002034 00:00:00:00:00:00 > 00:00:00:00:00:00, ethertype IPv4 (0x0800), length 66: 127.0.0.1.9557 > 127.0.0.1.36828: Flags [.], ack 3562800815, win 15342, options [nop,nop,TS val 843174495 ecr 843159493], length 0
03:09:05.002094 00:00:00:00:00:00 > 00:00:00:00:00:00, ethertype IPv4 (0x0800), length 66: 127.0.0.1.36828 > 127.0.0.1.9557: Flags [.], ack 1, win 6160, options [nop,nop,TS val 843174496 ecr 843159493], length 0
2 packets captured
116 packets received by filter
0 packets dropped by kernel
```
Wireshark shows a similar result as the tcpdump example above.
Linux's loopback setup: https://github.com/torvalds/linux/blob/5bfc75d92efd494db37f5c4c173d3639d4772966/drivers/net/loopback.c#L162
PiperOrigin-RevId: 391836719
|
|
Also fix an option parsing error in checker.TCPTimestampChecker while I am here.
PiperOrigin-RevId: 391828329
|
|
Use different secrets for different purposes (port picking,
ISN generation, tsOffset generation) and moved the secrets
from stack.Stack to tcp.protocol.
PiperOrigin-RevId: 391641238
|
|
-Added a test to check if the process in background is blocking SIGTTOU
-Some minor formatting fixes
|
|
Also plumber the controls through runsc
PiperOrigin-RevId: 391594318
|
|
nogo_stdlib propogate facts and findings to downstream nogo_aspects via
NogoStdlibInfo. This all works fine except one case: directly building a
nogo_stdlib. e.g., bazel build //tools/nogo:stdlib.
In this case, nothing is requesting the NogoStdlibInfo, and thus the target has
nothing to do. This can be rather confusing when trying to debug failures in
:stdlib, as building :stdlib does nothing.
Fix this by declaring the facts and findings as default outputs from
nogo_stdlib. Now direct bazel build will request these outputs and actually
trigger the analysis. Standard aspect builds are unaffected.
PiperOrigin-RevId: 391580126
|
|
Earlier the tests were checking for equality of system-wide metrics before and
after some network related operations. That is inherently racy for native tests
because depending on the testing infrastructure, multiple tests might run
parallely hence trampling over each other's metrics.
Tests should only compare metrics that are increasing in nature. The comparison
should not be a hard comparison, instead a less-than/greater-than relation test.
I have changed the checks and also removed tests for tcpCurrEstab metric which
has "SYNTAX Gauge" and hence can not be tested reliably.
PiperOrigin-RevId: 391460081
|
|
PiperOrigin-RevId: 391416650
|
|
|
|
Test the correct sending of the SIGTTOU in setForegroundProcess
|
|
The test expects that pread reads the full buffer, it means that the pread
offset has to be equal or less than file_size - buffer_size.
PiperOrigin-RevId: 391356863
|
|
Add support for msgctl and enable tests.
Fixes #135
|
|
For IPCInfo, update value of MSGSEG constant in abi to avoid overflow in
MsgInfo.MsgSeg. MSGSEG was originaly simplified in abi, and is unused
(by us and within the kernel), so updating it is okay.
Updates #135
|
|
Set provides functionality of {sem,shm,msg}ctl(IPC_SET).
|
|
Updates #135
|
|
PiperOrigin-RevId: 391331401
|
|
PiperOrigin-RevId: 391217339
|
|
This is a new kernel feature that are controlled by the net.core.mibs_allocation
sysctl.
PiperOrigin-RevId: 391215784
|
|
Otherwise, it fails with this error:
Get:3 http://security.debian.org/debian-security buster/updates InRelease
Reading package lists...
E: Repository 'http://deb.debian.org/debian buster InRelease' changed its
'Suite' value from 'stable' to 'oldstable'
PiperOrigin-RevId: 391155532
|
|
Changed sendSignal to sendSignalLocked because tg.pidns.owner.mu and tg.signalHandlers.mu
are already locked in SetForegroundProcess
Added a control to verify whether the calling process is ignoring SIGTTOU
before sending the signal
|
|
Removes package syserror and moves still relevant code to either linuxerr
or to syserr (to be later removed).
Internal errors are converted from random types to *errors.Error types used
in linuxerr. Internal errors are in linuxerr/internal.go.
PiperOrigin-RevId: 390724202
|
|
Update the start benchmark on empty to only "Start" a container, not wait
for its completion.
TL:DR only measure the actual start call for the empty container.
Previously, we were measuring the completion of /bin/true in container
alpine AND the cleanup. This was fine until profiling started failing all
the time on ptrace. This is a cost that runc is not paying.
These changes will reduce the over all timing of the benchmark, but it will
give more sane results.
Instead, use "Spawn" which is similar to `docker run --detach alpine
/bin/sleep 100`. Call sleep so containers stick around long enough
for the profiler to read profile data from them.
PiperOrigin-RevId: 390705431
|
|
Add Event controls and implement "stream" commands.
PiperOrigin-RevId: 390691702
|
|
PiperOrigin-RevId: 390659097
|
|
On Go tip (pre-1.18), http://golang.org/issue/44195 is making SA1019 mistake
uses of reflect.Value.Len for reflect.Value.InterfaceData, which is deprecated.
It is thus mistakenly raising deprecation errors on uses of reflect.Value.Len.
Suppress these errors by disabling SA1019 entirely. This is a bit overkill, but
it is unclear to me if we want hard errors on deprecation anyways. That can be
reevaluated when http://golang.org/issue/44195 is fixed.
The other staticcheck analyzers are moved to alphabetical order.
Updates golang/go#44195
PiperOrigin-RevId: 390655918
|
|
This list has gotten a little out-of-date. Note that `clockwork` used to be
used but was removed in gvisor.dev/pr/5384.
PiperOrigin-RevId: 390644841
|
|
tcpip.Endpoint.Close is documented to free all resources associated
with an endpoint so we don't need to create an empty map to clear
the multicast memberships.
PiperOrigin-RevId: 390609826
|
|
Add Usage controls and implement "usage/usagefd" commands.
PiperOrigin-RevId: 390507423
|
|
Convert remaining public errors (e.g. EINTR) from syserror to linuxerr.
PiperOrigin-RevId: 390471763
|