Files
e2787d3981 enhance: standardize error handling on merr + Sys/Input classification (#50221)
issue: #47420

## What this PR does

Project-wide migration of raw `fmt.Errorf` / `errors.New` in function
bodies onto
the `merr` framework, plus the Sys-vs-Input error classification and the
machinery it drives (retriability, fine-grained metrics, segcore
unification),
plus the convention docs and a linter that keeps it from regressing.

Scope: storage, proxy, coordinators (root/data/query), query node, data
node,
`pkg/util` & `internal/util`, expression parser, message queue,
streaming, and
misc packages. Bare raw-error usages went from ~3000 to a ~340 allowlist
(package-level sentinels / build-tag / test sites).

---

## How to review this PR

It is large but the vast majority is mechanical. Changes fall into three
tiers;
spend review budget on Part 2 and Part 3.

### Part 1 — Mechanical standardization (low risk, verify by rule)

Each converted call follows one of a small fixed set of rules. To
review, check
that each site obeys the matching rule rather than reading every line:

| Pattern | Rule |
|---|---|
| `fmt.Errorf("...")` originating a new error | →
`merr.WrapErrXxxMsg("...")` with a code matching the failure's meaning |
| Adding context to an existing typed error | → `merr.Wrap(err, "...")`
/ `merr.Wrapf(...)` — **preserves** the inner code (never `WrapErr*Err`,
which overwrites it) |
| Errors inside the streaming subsystem | → `status.New*` factories
(StreamingError), **not** merr — this is the component-internal dialect
(see `docs/dev/error_handling_guide.md`) |
| Low-level / control-flow signal caught by `errors.Is` | → kept as a
package-level `errors.New` sentinel (lowercase, same-package) |

Conventions are documented in `docs/dev/error_handling_guide.md`
(how-to) and
`docs/dev/error_sentinel_convention.md` (rules + audit). A
`gocritic`/`ruleguard`
rule (`rawmerrerror`, in `rules.go`) enforces "no raw `return
errors.New/fmt.Errorf`"
under `make verifiers`.

### Part 2 — Behavior changes (review these closely)

These are the sites where the wire contract or runtime behavior changes,
not just
the source text. Listed by category; representative locations given,
full set in
the diff.

**A. gRPC wire-code shifts: `UnexpectedError(1)/Code 65535` → typed
code.**
Where a handler previously returned a raw error (collapsed to
`Code=65535` on the
wire), it now returns a typed merr, so the client sees a real code. The
most
common shift is to `IllegalArgument(5)/Code 1100` (ParameterInvalid).
Touch
points include datanode task handlers (CreateTask/Query/Drop), proxy
Upsert,
querynode GetMetrics, datacoord CreateIndex, httpserver query-response
builder,
and typeutil schema validation. One code refinement: an index-param
validation
moved `1100` → `1101` (ParameterMissing). **Client/SDK assertions and
any code
that switched on `Code=65535` for these paths must be re-checked** (the
go_client
e2e assertions were already aligned in this PR).

**B. Prometheus `status` label contract change (externally visible).**
The proxy metric's coarse `fail` / `rejected` values are split into
`fail_input` / `fail_system` and `rejected_user` / `rejected_system` (in
`requestutil.ParseMetricLabel`; auth/privilege rejections count as
`rejected_user`), so dashboards can attribute a failure to caller vs
operator.
**Dashboards/alerts querying `status="fail"` must migrate to
`status=~"fail_.*"`, and `status="rejected"` to
`status=~"rejected_.*"`.** The
in-repo Grafana dashboard is already migrated; external dashboards built
on the
old values silently go empty after upgrade. This is the one change that
requires an ops-side migration.

**C. Retriability semantics.**
- C1: `merr.Status(err)` now forces `Retriable=false` when the error is
an
`InputError` — a malformed request can never succeed on blind retry, so
clients
never get the self-contradictory "your input is wrong but you may
retry".
- C2: `retry.Do` short-circuits an `InputError` (non-retriable) — **but
only when
  the caller did not pass a `RetryErr` predicate**. The check is an
`if c.isRetryErr != nil { ... } else if InputError { ... }` *mutually
exclusive*
branch (`pkg/util/retry/retry.go`): an explicit `RetryErr` takes
precedence and
bypasses the InputError abort. `retry.Handle` deliberately does **not**
apply
the InputError abort (its callers signal abort via `shouldRetry=false`).
Four
flusher startup callsites that must retry through transient "not ready"
errors
  were given explicit `RetryErr` escape hatches.

**D. segcore (C++→Go) error classification.**
A single shared Go-side table (`pkg/util/merr/segcore.go`) maps each
segcore code
to a merr sentinel + InputError/signal category, replacing scattered
hand-written
`if errorCode == ...` switches in the cgo wrappers. **Wire `Code` values
change
for every segcore pass-through error, not just the remapped ones.**
Named
sentinels remap (C++ `2003` → merr `2001`, `2033` → `2002`,
Folly/Knowhere codes
likewise); **all remaining pass-through codes (`2004`–`2043`, previously
surfaced to clients as raw C++ enum values) now serialize as `2000`**
(`ErrSegcore`), with the original C++ code preserved in the `Reason`
text
(`segcoreCode=...`); unknown/future codes collapse to `2000` as well
(pinned by
the `wire_code_projection` test). Transient segcore classes (object
storage /
file IO / OOM / mmap / FieldNotLoaded — 11 codes) now report
`Retriable=true`.
**Any client switching on raw segcore codes in the `2004`–`2043` range
must be
re-checked**; the in-Reason code remains available for diagnostics.
Signal
codes (PretendFinished / FollyCancel) are recognized centrally.
`errors.Is`-based
control flow on these (e.g. scheduler skip/retry) is preserved.

**E. InputError classification (25 sentinels + dynamic marks).**
25 sentinels in `errors.go` carry `WithErrorType(InputError)` (the
Collection /
ResourceGroup / Database families, `ErrIndexDuplicate`,
`ErrParameterInvalid`,
`ErrPrivilegeNotAuthenticated`, `ErrImportFailed`, `ErrQueryPlan`, ...),
plus dynamic
marks for the 8 segcore input codes (ExprInvalid, DimNotMatch,
MetricTypeInvalid, FieldIDInvalid, ...) and
`WrapErrAsInputError`. The widest blast radius is `ErrParameterInvalid`
(1100):
~2335 `WrapErrParameterInvalid*` callsites now classify as input /
non-retriable. Because of C1/C2 this changes retriability for
any path that returns these. **The audit to confirm no transient path
was
mis-marked is the single most important review item** (see Part 3). One
reverse
correction: storage field-stats parsing moved from `ErrParameterInvalid`
(input)
to `ErrDataIntegrity` — a corrupted stored stat is data corruption, not
user
input.

### Part 3 — Known risks & traps (called out proactively)

1. **`merr.Wrap` vs `WrapErr*Err` (code-masking).** `WrapErr*Err` builds
a
`wrappedMilvusError{sentinel: ErrServiceInternal}` whose `code()`
returns the
*outer* sentinel — it overwrites the inner typed code and hides the
`errors.Is`
chain. This is intentional (use it to *deliberately* downgrade), but it
was a
recurring conversion defect; the rule "add context with `merr.Wrap`,
downgrade
with `WrapErr*Err`" is enforced by convention and reviewed across the
diff.
2. **InputError × `retry.Do` blast radius.** Marking a sentinel
`InputError` makes
any `retry.Do(...)` without a `RetryErr` predicate stop retrying it.
Reviewers
should sanity-check that no transient use of the 19 newly-marked
sentinels
(especially `ErrParameterInvalid`) sits inside a retry loop that needed
to keep
   spinning. The known flusher cases were handled (see C2).
3. **The ~340 raw-error allowlist.** What remains as bare `errors.New`
is, by
design: package-level sentinels (caught by `errors.Is`), `//go:build
test`
sites, and out-of-band trees (`cmd/`, `tests/`, codegen, walimpls). The
linter
only bans the *direct-return* form; assignment-then-return escapes and
the full
no-exceptions ban are deferred to an AST-based linter (Tier 2,
documented).
4. **segcore C++ second step deferred.** This PR unifies classification
on the Go
side; splitting the dual-semantic C++ codes at the source is a
follow-up.

---

## Validation

- `make verifiers`: Go side clean (gofmt + static-check across modules,
including
  the new `rawmerrerror` rule with a 0-hit baseline repo-wide).
- `make test-go`: passing; the one real regression introduced (a
datanode
`invalid_task_type` assertion shifting `1` → `5` from a ParameterInvalid
  conversion) was fixed in-tree.
- go_client e2e CreateIndex assertions aligned to the new merr messages.

---------

Signed-off-by: zhenshan.cao <zhenshan.cao@zilliz.com>
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-06-12 15:04:51 -07:00

480 lines
17 KiB
Go

// Licensed to the LF AI & Data foundation under one
// or more contributor license agreements. See the NOTICE file
// distributed with this work for additional information
// regarding copyright ownership. The ASF licenses this file
// to you under the Apache License, Version 2.0 (the
// "License"); you may not use this file except in compliance
// with the License. You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package gorules
import (
"github.com/quasilyte/go-ruleguard/dsl"
)
// This is a collection of rules for ruleguard: https://github.com/quasilyte/go-ruleguard
// Remove extra conversions: mdempsky/unconvert
func unconvert(m dsl.Matcher) {
m.Match("int($x)").Where(m["x"].Type.Is("int") && !m["x"].Const).Report("unnecessary conversion").Suggest("$x")
m.Match("float32($x)").Where(m["x"].Type.Is("float32") && !m["x"].Const).Report("unnecessary conversion").Suggest("$x")
m.Match("float64($x)").Where(m["x"].Type.Is("float64") && !m["x"].Const).Report("unnecessary conversion").Suggest("$x")
// m.Match("byte($x)").Where(m["x"].Type.Is("byte")).Report("unnecessary conversion").Suggest("$x")
// m.Match("rune($x)").Where(m["x"].Type.Is("rune")).Report("unnecessary conversion").Suggest("$x")
m.Match("bool($x)").Where(m["x"].Type.Is("bool") && !m["x"].Const).Report("unnecessary conversion").Suggest("$x")
m.Match("int8($x)").Where(m["x"].Type.Is("int8") && !m["x"].Const).Report("unnecessary conversion").Suggest("$x")
m.Match("int16($x)").Where(m["x"].Type.Is("int16") && !m["x"].Const).Report("unnecessary conversion").Suggest("$x")
m.Match("int32($x)").Where(m["x"].Type.Is("int32") && !m["x"].Const).Report("unnecessary conversion").Suggest("$x")
m.Match("int64($x)").Where(m["x"].Type.Is("int64") && !m["x"].Const).Report("unnecessary conversion").Suggest("$x")
m.Match("uint8($x)").Where(m["x"].Type.Is("uint8") && !m["x"].Const).Report("unnecessary conversion").Suggest("$x")
m.Match("uint16($x)").Where(m["x"].Type.Is("uint16") && !m["x"].Const).Report("unnecessary conversion").Suggest("$x")
m.Match("uint32($x)").Where(m["x"].Type.Is("uint32") && !m["x"].Const).Report("unnecessary conversion").Suggest("$x")
m.Match("uint64($x)").Where(m["x"].Type.Is("uint64") && !m["x"].Const).Report("unnecessary conversion").Suggest("$x")
m.Match("time.Duration($x)").Where(m["x"].Type.Is("time.Duration") && !m["x"].Text.Matches("^[0-9]*$")).Report("unnecessary conversion").Suggest("$x")
}
// Don't use == or != with time.Time
// https://github.com/dominikh/go-tools/issues/47 : Wontfix
func timeeq(m dsl.Matcher) {
m.Match("$t0 == $t1").Where(m["t0"].Type.Is("time.Time")).Report("using == with time.Time")
m.Match("$t0 != $t1").Where(m["t0"].Type.Is("time.Time")).Report("using != with time.Time")
m.Match(`map[$k]$v`).Where(m["k"].Type.Is("time.Time")).Report("map with time.Time keys are easy to misuse")
}
// err but no an error
func errnoterror(m dsl.Matcher) {
// Would be easier to check for all err identifiers instead, but then how do we get the type from m[] ?
m.Match(
"if $*_, err := $x; $err != nil { $*_ } else if $_ { $*_ }",
"if $*_, err := $x; $err != nil { $*_ } else { $*_ }",
"if $*_, err := $x; $err != nil { $*_ }",
"if $*_, err = $x; $err != nil { $*_ } else if $_ { $*_ }",
"if $*_, err = $x; $err != nil { $*_ } else { $*_ }",
"if $*_, err = $x; $err != nil { $*_ }",
"$*_, err := $x; if $err != nil { $*_ } else if $_ { $*_ }",
"$*_, err := $x; if $err != nil { $*_ } else { $*_ }",
"$*_, err := $x; if $err != nil { $*_ }",
"$*_, err = $x; if $err != nil { $*_ } else if $_ { $*_ }",
"$*_, err = $x; if $err != nil { $*_ } else { $*_ }",
"$*_, err = $x; if $err != nil { $*_ }",
).
Where(m["err"].Text == "err" && !m["err"].Type.Is("error") && m["x"].Text != "recover()").
Report("err variable not error type")
}
// Identical if and else bodies
func ifbodythenbody(m dsl.Matcher) {
m.Match("if $*_ { $body } else { $body }").
Report("identical if and else bodies")
// Lots of false positives.
// m.Match("if $*_ { $body } else if $*_ { $body }").
// Report("identical if and else bodies")
}
// Odd inequality: A - B < 0 instead of !=
// Too many false positives.
/*
func subtractnoteq(m dsl.Matcher) {
m.Match("$a - $b < 0").Report("consider $a != $b")
m.Match("$a - $b > 0").Report("consider $a != $b")
m.Match("0 < $a - $b").Report("consider $a != $b")
m.Match("0 > $a - $b").Report("consider $a != $b")
}
*/
// Self-assignment
func selfassign(m dsl.Matcher) {
m.Match("$x = $x").Report("useless self-assignment")
}
// Odd nested ifs
func oddnestedif(m dsl.Matcher) {
m.Match("if $x { if $x { $*_ }; $*_ }",
"if $x == $y { if $x != $y {$*_ }; $*_ }",
"if $x != $y { if $x == $y {$*_ }; $*_ }",
"if $x { if !$x { $*_ }; $*_ }",
"if !$x { if $x { $*_ }; $*_ }").
Report("odd nested ifs")
m.Match("for $x { if $x { $*_ }; $*_ }",
"for $x == $y { if $x != $y {$*_ }; $*_ }",
"for $x != $y { if $x == $y {$*_ }; $*_ }",
"for $x { if !$x { $*_ }; $*_ }",
"for !$x { if $x { $*_ }; $*_ }").
Report("odd nested for/ifs")
}
// odd bitwise expressions
func oddbitwise(m dsl.Matcher) {
m.Match("$x | $x",
"$x | ^$x",
"^$x | $x").
Report("odd bitwise OR")
m.Match("$x & $x",
"$x & ^$x",
"^$x & $x").
Report("odd bitwise AND")
m.Match("$x &^ $x").
Report("odd bitwise AND-NOT")
}
// odd sequence of if tests with return
func ifreturn(m dsl.Matcher) {
m.Match("if $x { return $*_ }; if $x {$*_ }").Report("odd sequence of if test")
m.Match("if $x { return $*_ }; if !$x {$*_ }").Report("odd sequence of if test")
m.Match("if !$x { return $*_ }; if $x {$*_ }").Report("odd sequence of if test")
m.Match("if $x == $y { return $*_ }; if $x != $y {$*_ }").Report("odd sequence of if test")
m.Match("if $x != $y { return $*_ }; if $x == $y {$*_ }").Report("odd sequence of if test")
}
func oddifsequence(m dsl.Matcher) {
/*
m.Match("if $x { $*_ }; if $x {$*_ }").Report("odd sequence of if test")
m.Match("if $x == $y { $*_ }; if $y == $x {$*_ }").Report("odd sequence of if tests")
m.Match("if $x != $y { $*_ }; if $y != $x {$*_ }").Report("odd sequence of if tests")
m.Match("if $x < $y { $*_ }; if $y > $x {$*_ }").Report("odd sequence of if tests")
m.Match("if $x <= $y { $*_ }; if $y >= $x {$*_ }").Report("odd sequence of if tests")
m.Match("if $x > $y { $*_ }; if $y < $x {$*_ }").Report("odd sequence of if tests")
m.Match("if $x >= $y { $*_ }; if $y <= $x {$*_ }").Report("odd sequence of if tests")
*/
}
// odd sequence of nested if tests
func nestedifsequence(m dsl.Matcher) {
/*
m.Match("if $x < $y { if $x >= $y {$*_ }; $*_ }").Report("odd sequence of nested if tests")
m.Match("if $x <= $y { if $x > $y {$*_ }; $*_ }").Report("odd sequence of nested if tests")
m.Match("if $x > $y { if $x <= $y {$*_ }; $*_ }").Report("odd sequence of nested if tests")
m.Match("if $x >= $y { if $x < $y {$*_ }; $*_ }").Report("odd sequence of nested if tests")
*/
}
// odd sequence of assignments
func identicalassignments(m dsl.Matcher) {
m.Match("$x = $y; $y = $x").Report("odd sequence of assignments")
}
func oddcompoundop(m dsl.Matcher) {
m.Match("$x += $x + $_",
"$x += $x - $_").
Report("odd += expression")
m.Match("$x -= $x + $_",
"$x -= $x - $_").
Report("odd -= expression")
}
func constswitch(m dsl.Matcher) {
m.Match("switch $x { $*_ }", "switch $*_; $x { $*_ }").
Where(m["x"].Const && !m["x"].Text.Matches(`^runtime\.`)).
Report("constant switch")
}
func oddcomparisons(m dsl.Matcher) {
m.Match(
"$x - $y == 0",
"$x - $y != 0",
"$x - $y < 0",
"$x - $y <= 0",
"$x - $y > 0",
"$x - $y >= 0",
"$x ^ $y == 0",
"$x ^ $y != 0",
).Report("odd comparison")
}
func oddmathbits(m dsl.Matcher) {
m.Match(
"64 - bits.LeadingZeros64($x)",
"32 - bits.LeadingZeros32($x)",
"16 - bits.LeadingZeros16($x)",
"8 - bits.LeadingZeros8($x)",
).Report("odd math/bits expression: use bits.Len*() instead?")
}
// func floateq(m dsl.Matcher) {
// m.Match(
// "$x == $y",
// "$x != $y",
// ).
// Where(m["x"].Type.Is("float32") && !m["x"].Const && !m["y"].Text.Matches("0(.0+)?") && !m.File().Name.Matches("floating_comparision.go")).
// Report("floating point tested for equality")
// m.Match(
// "$x == $y",
// "$x != $y",
// ).
// Where(m["x"].Type.Is("float64") && !m["x"].Const && !m["y"].Text.Matches("0(.0+)?") && !m.File().Name.Matches("floating_comparision.go")).
// Report("floating point tested for equality")
// m.Match("switch $x { $*_ }", "switch $*_; $x { $*_ }").
// Where(m["x"].Type.Is("float32")).
// Report("floating point as switch expression")
// m.Match("switch $x { $*_ }", "switch $*_; $x { $*_ }").
// Where(m["x"].Type.Is("float64")).
// Report("floating point as switch expression")
// }
func badexponent(m dsl.Matcher) {
m.Match(
"2 ^ $x",
"10 ^ $x",
).
Report("caret (^) is not exponentiation")
}
func floatloop(m dsl.Matcher) {
m.Match(
"for $i := $x; $i < $y; $i += $z { $*_ }",
"for $i = $x; $i < $y; $i += $z { $*_ }",
).
Where(m["i"].Type.Is("float64")).
Report("floating point for loop counter")
m.Match(
"for $i := $x; $i < $y; $i += $z { $*_ }",
"for $i = $x; $i < $y; $i += $z { $*_ }",
).
Where(m["i"].Type.Is("float32")).
Report("floating point for loop counter")
}
func urlredacted(m dsl.Matcher) {
m.Match(
"log.Println($x, $*_)",
"log.Println($*_, $x, $*_)",
"log.Println($*_, $x)",
"log.Printf($*_, $x, $*_)",
"log.Printf($*_, $x)",
"log.Println($x, $*_)",
"log.Println($*_, $x, $*_)",
"log.Println($*_, $x)",
"log.Printf($*_, $x, $*_)",
"log.Printf($*_, $x)",
).
Where(m["x"].Type.Is("*url.URL")).
Report("consider $x.Redacted() when outputting URLs")
}
func sprinterr(m dsl.Matcher) {
m.Match(`fmt.Sprint($err)`,
`fmt.Sprintf("%s", $err)`,
`fmt.Sprintf("%v", $err)`,
).
Where(m["err"].Type.Is("error")).
Report("maybe call $err.Error() instead of fmt.Sprint()?")
}
// disable this check, because it can not apply to generic type
//func largeloopcopy(m dsl.Matcher) {
// m.Match(
// `for $_, $v := range $_ { $*_ }`,
// ).
// Where(m["v"].Type.Size > 1024).
// Report(`loop copies large value each iteration`)
//}
func joinpath(m dsl.Matcher) {
m.Match(
`strings.Join($_, "/")`,
`strings.Join($_, "\\")`,
"strings.Join($_, `\\`)",
).
Report(`did you mean path.Join() or filepath.Join() ?`)
}
func readfull(m dsl.Matcher) {
m.Match(`$n, $err := io.ReadFull($_, $slice)
if $err != nil || $n != len($slice) {
$*_
}`,
`$n, $err := io.ReadFull($_, $slice)
if $n != len($slice) || $err != nil {
$*_
}`,
`$n, $err = io.ReadFull($_, $slice)
if $err != nil || $n != len($slice) {
$*_
}`,
`$n, $err = io.ReadFull($_, $slice)
if $n != len($slice) || $err != nil {
$*_
}`,
`if $n, $err := io.ReadFull($_, $slice); $n != len($slice) || $err != nil {
$*_
}`,
`if $n, $err := io.ReadFull($_, $slice); $err != nil || $n != len($slice) {
$*_
}`,
`if $n, $err = io.ReadFull($_, $slice); $n != len($slice) || $err != nil {
$*_
}`,
`if $n, $err = io.ReadFull($_, $slice); $err != nil || $n != len($slice) {
$*_
}`,
).Report("io.ReadFull() returns err == nil iff n == len(slice)")
}
func nilerr(m dsl.Matcher) {
m.Match(
`if err == nil { return err }`,
`if err == nil { return $*_, err }`,
).
Report(`return nil error instead of nil value`)
}
func mailaddress(m dsl.Matcher) {
m.Match(
"fmt.Sprintf(`\"%s\" <%s>`, $NAME, $EMAIL)",
"fmt.Sprintf(`\"%s\"<%s>`, $NAME, $EMAIL)",
"fmt.Sprintf(`%s <%s>`, $NAME, $EMAIL)",
"fmt.Sprintf(`%s<%s>`, $NAME, $EMAIL)",
`fmt.Sprintf("\"%s\"<%s>", $NAME, $EMAIL)`,
`fmt.Sprintf("\"%s\" <%s>", $NAME, $EMAIL)`,
`fmt.Sprintf("%s<%s>", $NAME, $EMAIL)`,
`fmt.Sprintf("%s <%s>", $NAME, $EMAIL)`,
).
Report("use net/mail Address.String() instead of fmt.Sprintf()").
Suggest("(&mail.Address{Name:$NAME, Address:$EMAIL}).String()")
}
func errnetclosed(m dsl.Matcher) {
m.Match(
`strings.Contains($err.Error(), $text)`,
).
Where(m["text"].Text.Matches("\".*closed network connection.*\"")).
Report(`String matching against error texts is fragile; use net.ErrClosed instead`).
Suggest(`errors.Is($err, net.ErrClosed)`)
}
func httpheaderadd(m dsl.Matcher) {
m.Match(
`$H.Add($KEY, $VALUE)`,
).
Where(m["H"].Type.Is("http.Header")).
Report("use http.Header.Set method instead of Add to overwrite all existing header values").
Suggest(`$H.Set($KEY, $VALUE)`)
}
func hmacnew(m dsl.Matcher) {
m.Match("hmac.New(func() hash.Hash { return $x }, $_)",
`$f := func() hash.Hash { return $x }
$*_
hmac.New($f, $_)`,
).Where(m["x"].Pure).
Report("invalid hash passed to hmac.New()")
}
func writestring(m dsl.Matcher) {
m.Match(`io.WriteString($w, string($b))`).
Where(m["b"].Type.Is("[]byte")).
Suggest("$w.Write($b)")
}
func badlock(m dsl.Matcher) {
// Shouldn't give many false positives without type filter
// as Lock+Unlock pairs in combination with defer gives us pretty
// a good chance to guess correctly. If we constrain the type to sync.Mutex
// then it'll be harder to match embedded locks and custom methods
// that may forward the call to the sync.Mutex (or other synchronization primitive).
m.Match(`$mu.Lock(); defer $mu.RUnlock()`).Report(`maybe $mu.RLock() was intended?`)
m.Match(`$mu.RLock(); defer $mu.Unlock()`).Report(`maybe $mu.Lock() was intended?`)
}
// Raw error constructors (fmt.Errorf / errors.New / errors.Errorf) must not be
// returned directly from a function body. An error that reaches the gRPC
// boundary has to carry a merr code, so originate it through the merr framework
// (merr.WrapErr*Msg / merr.WrapErr*Err) and add context with merr.Wrap/Wrapf.
//
// Scope: this catches the direct-return form, the form that lets a raw error
// escape to the boundary. Sentinels are expected to be package-level
// (var Err... = errors.New(...)) — ValueSpec nodes that the statement patterns
// below never match — so they are exempt automatically; function-local
// sentinels should be lifted to package level rather than kept inline.
// Assignment-then-return escapes (e := errors.New(); return e) cannot be
// expressed as a ruleguard pattern and are backstopped at the gRPC boundary by
// merr.Status. Test files, the cmd/ and tests/ trees, codegen and the walimpls
// harness run outside the request path and are exempt.
func rawmerrerror(m dsl.Matcher) {
// fmt.Errorf resolves unambiguously to the stdlib, so the literal selector matches.
m.Match(
`return fmt.Errorf($*_)`,
`return $*_, fmt.Errorf($*_)`,
`return fmt.Errorf($*_), $*_`,
).
Where(!m.File().Name.Matches(`_test\.go$`) &&
!m.File().Name.Matches(`test_streaming\.go$`) &&
!m.File().PkgPath.Matches(`/milvus/cmd/`) &&
!m.File().PkgPath.Matches(`/milvus/tests/`) &&
!m.File().PkgPath.Matches(`codegen`) &&
!m.File().PkgPath.Matches(`walimplstest`) &&
!m.File().PkgPath.Matches(`/mocks/`)).
Report(`raw error returned from function body; originate through the merr framework (merr.WrapErr*/merr.Wrap) instead of fmt.Errorf/errors.New/errors.Errorf`)
// errors.New/Newf/Errorf: the repo uses github.com/cockroachdb/errors, not the
// stdlib, so a literal `errors.New` selector — which ruleguard type-resolves to
// the stdlib package — never matches. Match by the function name instead, limited
// to the raw constructors; errors.Wrap/Wrapf/Is/As/Cause are intentionally allowed.
m.Match(
`return errors.$fn($*_)`,
`return $*_, errors.$fn($*_)`,
`return errors.$fn($*_), $*_`,
).
Where(m["fn"].Text.Matches(`^(New|Newf|Errorf)$`) &&
!m.File().Name.Matches(`_test\.go$`) &&
!m.File().Name.Matches(`test_streaming\.go$`) &&
!m.File().PkgPath.Matches(`/milvus/cmd/`) &&
!m.File().PkgPath.Matches(`/milvus/tests/`) &&
!m.File().PkgPath.Matches(`codegen`) &&
!m.File().PkgPath.Matches(`walimplstest`) &&
!m.File().PkgPath.Matches(`/mocks/`)).
Report(`raw error returned from function body; originate through the merr framework (merr.WrapErr*/merr.Wrap) instead of fmt.Errorf/errors.New/errors.Errorf`)
}
// merrsentinel forbids parking merr-typed errors in variables that look like
// sentinels. milvusError.Is compares error codes only, so errors.Is against a
// merr-typed "sentinel" silently matches ANY error sharing the code — a guard
// like errors.Is(err, ErrIgnoredFoo) then treats unrelated internal failures
// as the special case (real incident: stale-meta / already-loaded guards
// acknowledged etcd write failures as success). Identity sentinels must stay
// plain errors.New; merr errors must be created at the return site.
func merrsentinel(m dsl.Matcher) {
m.Match(
`var $x = merr.$f($*_)`,
`var $x = merr.$v`,
).
Where(!m.File().Name.Matches(`_test\.go$`) &&
!m.File().PkgPath.Matches(`/milvus/cmd/`) &&
!m.File().PkgPath.Matches(`/milvus/tests/`) &&
!m.File().PkgPath.Matches(`/mocks/`) &&
!m.File().PkgPath.Matches(`util/merr`)).
Report(`don't store a merr error in a sentinel-like variable: errors.Is on merr compares codes, not identity, so any same-code error would match. Use errors.New for identity sentinels, or create the merr error at the return site`)
}