The current implementation doesn't handle s.freeindex at all, which
means it'll skip a whole bunch of recently-made allocations, because the
span may have not been swept yet.
Change-Id: I8c5e360f5927ffe7e9abb448b352a59875e31b02
Reviewed-on: https://go-review.googlesource.com/c/go/+/586996
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Cleanup all remaining trivial compares against $0 in ppc64x assembly.
In math, SRD ...,Rx; CMP Rx, $0 is further simplified to SRDCC.
Change-Id: Ia2bc204953e32f08ee142bfd06a91965f30f99b6
Reviewed-on: https://go-review.googlesource.com/c/go/+/587016
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Run-TryBot: Paul Murphy <murp@ibm.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Tests of the mutex profile focus on sync.Mutex, which is easy to
control. But since those tests still use the runtime, and contention on
internal runtime.mutex values is now also part of the mutex profile, we
have to filter out those samples before examining the profile. Otherwise
the test may be confused by stray contention on sched.lock (or other
runtime-internal locks) as a natural consequence of using goroutines.
Fixes#67563
Change-Id: I066a24674d8b719dbeca4a5c0f76b53bc07498c1
Reviewed-on: https://go-review.googlesource.com/c/go/+/586957
Reviewed-by: Cherry Mui <cherryyz@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Rhys Hiltner <rhys.hiltner@gmail.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
There is only one trusted certificate I could find in the web pki which
has a negative serial number. Removing this exception seems reasonable.
Updates #65085
Change-Id: I55435b3d75479dcb41d523383e4ff7894a1496ad
Reviewed-on: https://go-review.googlesource.com/c/go/+/562343
Auto-Submit: Roland Shoemaker <roland@golang.org>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Damien Neil <dneil@google.com>
Mutex contention measurements work with two clocks: nanotime for use in
runtime/metrics, and cputicks for the runtime/pprof profile. They're
subject to different sampling rates: the runtime/metrics view is always
enabled, but the profile is adjustable and is turned off by default.
They have different levels of overhead: it can take as little as one
instruction to read cputicks while nanotime calls are more elaborate
(although some platforms implement cputicks as a nanotime call). The use
of the timestamps is also different: the profile's view needs to attach
the delay in some Ms' lock2 calls to another M's unlock2 call stack, but
the metric's view is only an int64.
Treat them differently. Don't bother threading the nanotime clock
through to the unlock2 call, measure and report it directly within
lock2. Sample nanotime at a constant gTrackingPeriod.
Don't consult any clocks unless the mutex is actually contended.
Continue liberal use of cputicks for now.
For #66999
Change-Id: I1c2085ea0e695bfa90c30fadedc99ced9eb1f69e
Reviewed-on: https://go-review.googlesource.com/c/go/+/586796
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Rhys Hiltner <rhys.hiltner@gmail.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Run-TryBot: Rhys Hiltner <rhys.hiltner@gmail.com>
Reviewed-by: Carlos Amedee <carlos@golang.org>
Move code so that basic imports work instead
of //go:linkname for metadata lists.
For #67401.
Change-Id: Id02075570befc45a9426559aad2137ab540928b2
Reviewed-on: https://go-review.googlesource.com/c/go/+/585915
Reviewed-by: Than McIntosh <thanm@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Modify rangefunc #next protocol to make it more robust
Extra-terrible nests of rangefunc iterators caused the
prior implementation to misbehave non-locally (in outer loops).
Add more rangefunc exit flag tests, parallel and tricky
This tests the assertion that a rangefunc iterator running
in parallel can trigger the race detector if any of the
parallel goroutines attempts an early exit. It also
verifies that if everything else is carefully written,
that it does NOT trigger the race detector if all the
parts run time completion.
Another test tries to rerun a yield function within a loop,
so that any per-line shared checking would be fooled.
Added all the use-of-body/yield-function checking.
These checks handle pathological cases that would cause
rangefunc for loops to behave in surprising ways (compared
to "regular" for loops). For example, a rangefunc iterator
might defer-recover a panic thrown in the syntactic body
of a loop; this notices the fault and panics with an
explanation
Modified closure naming to ID rangefunc bodies
Add a "-range<N>" suffix to the name of any closure generated for
a rangefunc loop body, as provided in Alessandro Arzilli's CL
(which is merged into this one).
Fix return values for panicky range functions
This removes the delayed implementation of "return x" by
ensuring that return values (in rangefunc-return-containing
functions) always have names and translating the "return x"
into "#rv1 = x" where #rv1 is the synthesized name of the
first result.
Updates #61405.
Change-Id: I933299ecce04ceabcf1c0c2de8e610b2ecd1cfd8
Reviewed-on: https://go-review.googlesource.com/c/go/+/584596
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Tim King <taking@google.com>
Move code to internal/coverage/cfile, making it possible to
access directly from testing/internal/testdeps, so that we can
avoid needing //go:linkname hacks.
For #67401.
Change-Id: I10b23a9970164afd2165e718ef3b2d9e86783883
Reviewed-on: https://go-review.googlesource.com/c/go/+/585820
Auto-Submit: Russ Cox <rsc@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Have the test use the same clock (cputicks) as the profiler, and use the
test's own measurements as hard bounds on the magnitude to expect in the
profile.
Compare the depiction of two users of the same lock: one where the
critical section is fast, one where it is slow. Confirm that the profile
shows the slow critical section as a large source of delay (with #66999
fixed), rather than showing the fast critical section as a large
recipient of delay.
For #64253
For #66999
Change-Id: I784c8beedc39de564dc8cee42060a5d5ce55de39
Reviewed-on: https://go-review.googlesource.com/c/go/+/586237
Auto-Submit: Rhys Hiltner <rhys.hiltner@gmail.com>
Reviewed-by: Carlos Amedee <carlos@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Go 1.22 promised to remove the setting in a future release once the
semantics of runtime-internal lock contention matched that of
sync.Mutex. That work is done, remove the setting.
For #66999
Change-Id: I3c4894148385adf2756d8754e44d7317305ad758
Reviewed-on: https://go-review.googlesource.com/c/go/+/585639
Reviewed-by: Carlos Amedee <carlos@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Rhys Hiltner <rhys.hiltner@gmail.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
When an M's use of a lock causes delays in other Ms, capture the stack
of the unlock call that caused the delay. This makes the meaning of the
mutex profile for runtime-internal mutexes match the behavior for
sync.Mutex: the profile points to the end of the critical section that
is responsible for delaying other work.
Fixes#66999
Change-Id: I4abc4a1df00a48765d29c07776481a1cbd539ff8
Reviewed-on: https://go-review.googlesource.com/c/go/+/585638
Auto-Submit: Rhys Hiltner <rhys.hiltner@gmail.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Carlos Amedee <carlos@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
When an M unlocks a contended mutex, it needs to consult a list of the
Ms that had to wait during its critical section. This allows the M to
attribute the appropriate amount of blame to the unlocking call stack.
Mirroring the implementation for users' sync.Mutex contention (via
sudog), we can (in a future commit) use the time that the head and tail
of the wait list started waiting, and the number of waiters, to estimate
the sum of the Ms' delays.
When an M acquires the mutex, it needs to remove itself from the list of
waiters. Since the futex-based lock implementation leaves the OS in
control of the order of M wakeups, we need to be prepared for quickly
(constant time) removing any M from the list.
First, have each M add itself to a singly-linked wait list when it finds
that its lock call will need to sleep. This case is safe against
live-lock, since any delay to one M adding itself to the list would be
due to another M making durable progress.
Second, have the M that holds the lock (either right before releasing,
or right after acquiring) update metadata on the list of waiting Ms to
double-link the list and maintain a tail pointer and waiter count. That
work is amortized-constant: we'll avoid contended locks becoming
proportionally more contended and undergoing performance collapse.
For #66999
Change-Id: If75cdea915afb59ccec47294e0b52c466aac8736
Reviewed-on: https://go-review.googlesource.com/c/go/+/585637
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Rhys Hiltner <rhys.hiltner@gmail.com>
Prepare the futex-based implementation of lock2 to maintain a list of
waiting Ms. Beyond storing an muintptr in the mutex's key field, we now
must never overwrite that field (even for a moment) without taking its
current value into account.
The semaphore-based implementation of lock2 already has that behavior.
Reuse that structure.
For #66999
Change-Id: I23b6f6bacb276fe33c6aed5c0571161a7e71fe6c
Reviewed-on: https://go-review.googlesource.com/c/go/+/585636
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Rhys Hiltner <rhys.hiltner@gmail.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Move the nextwaitm field into a small struct, in preparation for
additional metadata to track how long Ms need to wait for locks.
For #66999
Change-Id: Ib40e43c15cde22f7e35922641107973d99439ecd
Reviewed-on: https://go-review.googlesource.com/c/go/+/585635
Reviewed-by: Michael Pratt <mpratt@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Rhys Hiltner <rhys.hiltner@gmail.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Allow users to decrease the profiling stack depth back to 32 in case
they experience any problems with the new default of 128.
Users may also use this option to increase the depth up to 1024.
Change-Id: Ieaab2513024915a223239278dd97a6e161dde1cf
Reviewed-on: https://go-review.googlesource.com/c/go/+/581917
Reviewed-by: Austin Clements <austin@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
The current stack depth limit for alloc, mutex, block, threadcreate and
goroutine profiles of 32 frequently leads to truncated stack traces in
production applications. Increase the limit to 128 which is the same
size used by the execution tracer.
Create internal/profilerecord to define variants of the runtime's
StackRecord, MemProfileRecord and BlockProfileRecord types that can hold
arbitrarily big stack traces. Implement internal profiling APIs based on
these new types and use them for creating protobuf profiles and to act
as shims for the public profiling APIs using the old types.
This will lead to an increase in memory usage for applications that
use the impacted profile types and have stack traces exceeding the
current limit of 32. Those applications will also experience a slight
increase in CPU usage, but this will hopefully soon be mitigated via CL
540476 and 533258 which introduce frame pointer unwinding for the
relevant profile types.
For #43669.
Change-Id: Ie53762e65d0f6295f5d4c7d3c87172d5a052164e
Reviewed-on: https://go-review.googlesource.com/c/go/+/572396
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Previously it was possible for mutex and block profile stack traces to
contain up to 32 frames in Stack0 or the resulting pprof profiles.
CL 533258 changed this behavior by using some of the space to
record skipped frames that are discarded when performing delayed inline
expansion. This has lowered the effective maximum stack size from 32 to
27 (the max skip value is 5), which can be seen as a small regression.
Add TestProfilerStackDepth to demonstrate the issue and protect all
profile types from similar regressions in the future. Fix the issue by
increasing the internal maxStack limit to take the maxSkip value into
account. Assert that the maxSkip value is never exceeded when recording
mutex and block profile stack traces.
Three alternative solutions to the problem were considered and
discarded:
1) Revert CL 533258 and give up on frame pointer unwinding. This seems
unappealing as we would lose the performance benefits of frame
pointer unwinding.
2) Discard skipped frames when recording the initial stack trace. This
would require eager inline expansion for up to maxSkip frames and
partially negate the performance benefits of frame pointer
unwinding.
3) Accept and document the new behavior. This would simplify the
implementation, but seems more confusing from a user perspective. It
also complicates the creation of test cases that make assertions
about the maximum profiling stack depth.
The execution tracer still has the same issue due to CL 463835. This
should be addressed in a follow-up CL.
Co-authored-by: Nick Ripley <nick.ripley@datadoghq.com>
Change-Id: Ibf4dbf08a5166c9cb32470068c69f58bc5f98d2c
Reviewed-on: https://go-review.googlesource.com/c/go/+/586657
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Change-Id: I9611edb9663a9c32bb23d26b9b22b79a54a97498
Reviewed-on: https://go-review.googlesource.com/c/go/+/586196
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
CL 544455, which added atomic And/Or APIs, raced with CL 585556, which
enabled stricter linkname checking. This caused linkname-related
failures on ARM and MIPS. Fix this by adding the necessary linknames.
We fix one other linkname that got overlooked in CL 585556.
Updates #61395.
Change-Id: I454f0767ce28188e550a61bc39b7e398239bc10e
Reviewed-on: https://go-review.googlesource.com/c/go/+/586516
Reviewed-by: Mauri de Souza Meneguzzo <mauri870@gmail.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Austin Clements <austin@google.com>
This change fixes problems with thread-locked goroutines using
newcoro/coroswitch/etc. Currently, the coro paths do not consider
thread-locked goroutines at all and can quickly result in broken
scheduler state or lost/leaked goroutines.
One possible fix to these issues is to fall back on goroutine+channel
semantics, but that turns out to be fairly complicated to implement and
results in significant performance cliffs. More complex thread-lock
state donation tricks also result in some fairly complicated state
tracking that doesn't seem worth it given the use-cases of iter.Pull
(and even then, there will be performance cliffs).
This change implements a much simpler, but more restrictive semantics.
In particular, thread-lock state is tied to the coro at the first call
to newcoro (i.e. iter.Pull). From then on, the invariant is that if the
coro has any thread-lock state *or* a goroutine calling into coroswitch
has any thread-lock state, that the full gamut of thread-lock state must
remain the same as it was when newcoro was called (the full gamut
meaning internal and external lock counts as well as the identity of the
thread that was locked to).
This semantics allows the common cases to be always fast, but comes with
a non-orthogonality caveat. Specifically, when iter.Pull is used in
conjunction with thread-locked goroutines, complex cases (passing next
between goroutines or passing yield between goroutines) are likely to
fail. Simple cases, where any number of iter.Pull iterators are used in
a straightforward way (nested, in series, etc.) from the same
goroutine, will work and will be guaranteed to be fast regardless of
thread-lock state.
This is a compromise for the near-term and we may consider lifting the
restrictions imposed by this CL in the future.
Fixes#65889.
Fixes#65946.
Change-Id: I3fb5791e36a61f5ded50226a229a79d28739b24e
Reviewed-on: https://go-review.googlesource.com/c/go/+/583675
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Austin Clements <austin@google.com>
This change moves the v2 tracer into the trace directory.
Updates #67367
Change-Id: I3657b4227002cb00fdf29c797434800ea796715e
Reviewed-on: https://go-review.googlesource.com/c/go/+/584538
Reviewed-by: Michael Knyszek <mknyszek@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This CL implements the new sync/atomic AND and OR apis as well as their race
counterparts.
Fixes#61395
Change-Id: I294eefe4b3ac27bc4ed237edcbfa88a8c646d86f
GitHub-Last-Rev: f174297007
GitHub-Pull-Request: golang/go#64331
Reviewed-on: https://go-review.googlesource.com/c/go/+/544455
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Auto-Submit: Austin Clements <austin@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Keith Randall <khr@google.com>
This change removes unused parts of the v1 tracer in preperation of
the move of the v2 tracer into the trace package.
Updates #67367
Change-Id: I3e53a8afdef72dc90c2d5b514380d1077d284bc7
Reviewed-on: https://go-review.googlesource.com/c/go/+/584537
Reviewed-by: Michael Knyszek <mknyszek@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
CL 585358 adds restrictions to disallow pull-only linknames
(currently off by default). Currently, there are quite some pull-
only linknames in user code in the wild. In order not to break
those, we add push linknames to allow them to be pulled. This CL
includes linknames found in a large code corpus (thanks Matthew
Dempsky and Michael Pratt for the analysis!), that are not
currently linknamed.
Updates #67401.
Change-Id: I32f5fc0c7a6abbd7a11359a025cfa2bf458fe767
Reviewed-on: https://go-review.googlesource.com/c/go/+/586137
Reviewed-by: Russ Cox <rsc@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This patch removes redundant guards for the crash stack feature, as all
supported Go architectures now have a crash stack implementation.
Change-Id: I7ffb61c57955778d687073418130b2aaab0ff183
GitHub-Last-Rev: 6ff56d6420
GitHub-Pull-Request: golang/go#67202
Reviewed-on: https://go-review.googlesource.com/c/go/+/583416
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Mauri de Souza Meneguzzo <mauri870@gmail.com>
callbackUpdateSystemStack contains a fast path to exit early without
update if SP is already within the g0.stack bounds.
This is not safe, as a subsequent call may have new stack bounds that
only partially overlap the old stack bounds. In this case it is possible
to see an SP that is in the old stack bounds, but very close to the
bottom of the bounds due to the partial overlap. In that case we're very
likely to "run out" of space on the system stack.
We only need to do this on extra Ms, as normal Ms have precise bounds
defined when we allocated the stack.
TSAN annotations are added to x_cgo_getstackbounds because bounds is a
pointer into the Go stack. The stack can be reused when an old thread
exits and a new thread starts, but TSAN can't see the synchronization
there. This isn't a new case, but we are now calling more often.
Fixes#62440.
Cq-Include-Trybots: luci.golang.try:gotip-linux-amd64-longtest
Change-Id: I5389050494987b7668d0b317fb92f85e61d798ac
Reviewed-on: https://go-review.googlesource.com/c/go/+/584597
Reviewed-by: Cherry Mui <cherryyz@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This is a placeholder for future options (e.g. JSON).
The parameter is temporarily variadic to avoid breaking
x/telemetry (see CL 585378), but I plan to remove
the "..." later this week.
Updates #67182
Change-Id: I3f6f39455d852f92902f8e3f007d3093cbe555db
Reviewed-on: https://go-review.googlesource.com/c/go/+/585557
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Austin Clements <austin@google.com>
As mentioned in CL 584598, linkname is a mechanism that, when
abused, can break API integrity and even safety of Go programs.
CL 584598 is a first step to restrict the use of linknames, by
implementing a blocklist. This CL takes a step further, tightening
up the restriction by allowing linkname references ("pull") only
when the definition side explicitly opts into it, by having a
linkname on the definition (possibly to itself). This way, it is at
least clear on the definition side that the symbol, despite being
unexported, is accessed outside of the package. Unexported symbols
without linkname can now be actually private. This is similar to
the symbol visibility rule used by gccgo for years (which defines
unexported non-linknamed symbols as C static symbols).
As there can be pull-only linknames in the wild that may be broken
by this change, we currently only enforce this rule for symbols
defined in the standard library. Push linknames are added in the
standard library to allow things build.
Linkname references to external (non-Go) symbols are still allowed,
as their visibility is controlled by the C symbol visibility rules
and enforced by the C (static or dynamic) linker.
Assembly symbols are treated similar to linknamed symbols.
This is controlled by -checklinkname linker flag, currently not
enabled by default. A follow-up CL will enable it by default.
Change-Id: I07344f5c7a02124dbbef0fbc8fec3b666a4b2b0e
Reviewed-on: https://go-review.googlesource.com/c/go/+/585358
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>
CL 581215 changed 'throw' so that instead of print(s) it called
a more complicated function, printpanicval, that statically
appeared to have convTstring in its call graph, even though this
isn't dynamically reachable when called with a string argument.
However, this caused the link-time static callgraph test to point
out that throw (which is called in nowritebarrierrec contexts
such as markgc) reaches a write barrier.
The solution is to inline and specialize the printpanicval
function for strings; it reduces to printindented.
Thanks to mpratt for pointing out that the reachability
check is on the fully lowered code, and is thus sensitive
to optimizations such as inlining.
I added an explanatory comment on the line that generates
the error message to help future users confused as I was.
Fixesgolang/go#67274
Change-Id: Ief110d554de365ce4c09509dceee000cbee30ad9
Reviewed-on: https://go-review.googlesource.com/c/go/+/584617
Reviewed-by: Than McIntosh <thanm@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Use frame pointer unwinding, where supported, to collect call stacks for
the block, and mutex profilers. This method of collecting call stacks is
typically an order of magnitude faster than callers/tracebackPCs. The
marginal benefit for these profile types is likely small compared to
using frame pointer unwinding for the execution tracer. However, the
block profiler can have noticeable overhead unless the sampling rate is
very high. Additionally, using frame pointer unwinding in more places
helps ensure more testing/support, which benefits systems like the
execution tracer which rely on frame pointer unwinding to be practical
to use.
Change-Id: I4b36c90cd2df844645fd275a41b247352d635727
Reviewed-on: https://go-review.googlesource.com/c/go/+/533258
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Auto-Submit: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
fpunwindExpand currently allocates a new slice to hold the expanded call
stack. In each place it's used, the resulting slice won't be needed
immediately afterward, so the allocation is wasteful. Refactor
fpunwindExpand to instead expand the call stack into a provided buffer.
Change-Id: I05b26c191a8f76404c21ccbe3bd422325540425b
Reviewed-on: https://go-review.googlesource.com/c/go/+/543715
Reviewed-by: Cherry Mui <cherryyz@google.com>
Auto-Submit: Cherry Mui <cherryyz@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Adds runtime.debugPinnerV1 which returns a runtime.Pinner object that
pins itself. This is intended to be used by debuggers in conjunction
with runtime.debugCall to keep heap memory reachable even if it isn't
referenced from anywhere else.
Change-Id: I508ee6a7b103e68df83c96f2e04a0599200300dc
Reviewed-on: https://go-review.googlesource.com/c/go/+/558276
Reviewed-by: Cherry Mui <cherryyz@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Austin Clements <austin@google.com>
Running with few threads usually does not need 500ms to crash, so let it
crash as soon as possible. While the test may caused more time on slow
machine, try to expand the sleep time in test.
Updates #64752
Change-Id: I635fab846bd5e1735808d4d47bb9032d5a04cc2b
GitHub-Last-Rev: 84f3844ac0
GitHub-Pull-Request: golang/go#65018
Reviewed-on: https://go-review.googlesource.com/c/go/+/554615
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Cherry Mui <cherryyz@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Temporarily mark the function runtime.throw as "go:noinline" for the
time being to work around problems introduced by CL 581215. We do not
ordinarily inline runtime.throw unless the build is beind done with an
elevated inline budget (e.g. "-gcflags=-l=4"), so this change should
only have an effect for those special builds.
Updates #67274.
Change-Id: I3811913b8d441e0ddb1d4c7d7297ef23555582a2
Reviewed-on: https://go-review.googlesource.com/c/go/+/584616
Reviewed-by: Alan Donovan <adonovan@google.com>
Auto-Submit: Than McIntosh <thanm@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Mauri de Souza Meneguzzo <mauri870@gmail.com>
This CL causes the printing of panic values to ensure that all
newlines in the output are immediately followed by a tab, so
that there is no way for a maliciously crafted panic value to
fool a program attempting to parse the traceback into thinking
that the panic value is in fact a goroutine stack.
See https://github.com/golang/go/issues/64590#issuecomment-1932675696
+ release note
Updates #64590
Updates #63455
Change-Id: I5142acb777383c0c122779d984e73879567dc627
Reviewed-on: https://go-review.googlesource.com/c/go/+/581215
Auto-Submit: Alan Donovan <adonovan@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Currently freeSpan is called before large object stats are updated when
sweeping large objects. This means heapStats.inHeap might get subtracted
before the large object is added to the largeFree field. The end result
is that the /memory/classes/heap/unused:bytes metric, which subtracts
live objects (alloc-free) from inHeap may overflow.
Fix this by always updating the large object stats before calling
freeSpan.
Fixes#67019.
Change-Id: Ib02bd8dcd1cf8cd1bc0110b6141e74f678c10445
Reviewed-on: https://go-review.googlesource.com/c/go/+/583380
Auto-Submit: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Felix Geisendörfer <felix.geisendoerfer@datadoghq.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
The page tracer's functionality is now captured by the regular execution
tracer as an experimental GODEBUG variable. This is a lot more usable
and maintainable than the page tracer, which is likely to have bitrotted
by this point. There's also no tooling available for the page tracer.
Change-Id: I2408394555e01dde75a522e9a489b7e55cf12c8e
Reviewed-on: https://go-review.googlesource.com/c/go/+/583379
Auto-Submit: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Carlos Amedee <carlos@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Move profiling pc buffers from being stack allocated to an m field.
This is motivated by the next patch, which will increase the default
stack depth to 128, which might lead to undesirable stack growth for
goroutines that produce profiling events.
Additionally, this change paves the way to make the stack depth
configurable via GODEBUG.
Change-Id: Ifa407f899188e2c7c0a81de92194fdb627cb4b36
Reviewed-on: https://go-review.googlesource.com/c/go/+/574699
Reviewed-by: Michael Knyszek <mknyszek@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
This change adds expensive alloc/free events to traces, guarded by a
GODEBUG that can be set at run time by mutating the GODEBUG environment
variable. This supersedes the alloc/free trace deleted in a previous CL.
There are two parts to this CL.
The first part is adding a mechanism for exposing experimental events
through the tracer and trace parser. This boils down to a new
ExperimentalEvent event type in the parser API which simply reveals the
raw event data for the event. Each experimental event can also be
associated with "experimental data" which is associated with a
particular generation. This experimental data is just exposed as a bag
of bytes that supplements the experimental events.
In the runtime, this CL organizes experimental events by experiment.
An experiment is defined by a set of experimental events and a single
special batch type. Batches of this special type are exposed through the
parser's API as the aforementioned "experimental data".
The second part of this CL is defining the AllocFree experiment, which
defines 9 new experimental events covering heap object alloc/frees, span
alloc/frees, and goroutine stack alloc/frees. It also generates special
batches that contain a type table: a mapping of IDs to type information.
Change-Id: I965c00e3dcfdf5570f365ff89d0f70d8aeca219c
Reviewed-on: https://go-review.googlesource.com/c/go/+/583377
Reviewed-by: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
allocfreetrace prints all allocations and frees to stderr. It's not
terribly useful because it has a really huge overhead, making it not
feasible to use except for the most trivial programs. A follow-up CL
will replace it with something that is both more thorough and also lower
overhead.
Change-Id: I1d668fee8b6aaef5251a5aea3054ec2444d75eb6
Reviewed-on: https://go-review.googlesource.com/c/go/+/583376
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Carlos Amedee <carlos@golang.org>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
If a signal lands on a non-Go thread, and Go code doesn't want to
handle it, currently we re-raise the signal in the signal handler
after uninstalling our handler, so the C code can handle it.
But if there is no C signal handler and the signal is ignored,
there is no need to re-raise the signal. Just ignore it. This
avoids uninstalling and reinstalling our handler, which, for some
reason, changes errno when TSAN is used. And TSAN does not like
errno being changed in the signal handler.
Not really sure if this is the bset of complete fix, but it does
fix the immediate problem, and it seems a reasonable thing to do
by itself.
Test case is CL 581722.
Fixes#66427.
Change-Id: I7a043d53059f1ff4080f4fc8ef4065d76ee7d78a
Reviewed-on: https://go-review.googlesource.com/c/go/+/582077
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This code was just missed during the cleanup. There's maybe some merit
to keeping OneNewExtraM, but it would still be fairly optimistic. It's
trivial to bring back, so delete it for now.
Change-Id: I2d033c6daae787e0e8d6b92524f3e59610e2599f
Reviewed-on: https://go-review.googlesource.com/c/go/+/583375
Reviewed-by: Carlos Amedee <carlos@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Michael Anthony Knyszek
Paul E. Murphy
Rhys Hiltner
Roland Shoemaker
Russ Cox
David Chase
Austin Clements
Felix Geisendörfer
Michael Mitchell
Carlos Amedee
Mauri de Souza Meneguzzo
Cherry Mui
Jes Cok
Alan Donovan
Michael Pratt
Nick Ripley
Guoqi Chen
qiulaidongfeng
Alessandro Arzilli
zzkcode
Than McIntosh