encoding/json: don't reset before returning buffer to pool

[andig]

Reset is already performed when retrieving from pool

[gopherbot]

This PR (HEAD: d9df74a) has been imported to Gerrit for code review.

Please visit https://go-review.googlesource.com/c/go/+/194338 to see it.

Tip: You can toggle comments from me using the comments slash command (e.g. /comments off)
See the Wiki page for more info

[gopherbot]

Message from Andreas Goetz:

Patch Set 1:

Kindly asking for review as you've done on c054747

---

Please don’t reply on this GitHub thread. Visit golang.org/cl/194338.
After addressing review feedback, remember to publish your drafts!

[gopherbot]

Message from Brad Fitzpatrick:

Patch Set 1: Run-TryBot+1

---

Please don’t reply on this GitHub thread. Visit golang.org/cl/194338.
After addressing review feedback, remember to publish your drafts!

[gopherbot]

Message from Gobot Gobot:

Patch Set 1:

TryBots beginning. Status page: https://farmer.golang.org/try?commit=62f068c3

---

Please don’t reply on this GitHub thread. Visit golang.org/cl/194338.
After addressing review feedback, remember to publish your drafts!

[gopherbot]

Message from Gobot Gobot:

Patch Set 1: TryBot-Result+1

TryBots are happy.

---

Please don’t reply on this GitHub thread. Visit golang.org/cl/194338.
After addressing review feedback, remember to publish your drafts!

[gopherbot]

Message from Brad Fitzpatrick:

Patch Set 1:

R=mvdan

This seems consistent with the other usage of the pool in stream.go.

---

Please don’t reply on this GitHub thread. Visit golang.org/cl/194338.
After addressing review feedback, remember to publish your drafts!

[gopherbot]

Message from Emmanuel Odeke:

Patch Set 1:

Decent benchmark numbers from this change in encoding/json:

name old time/op new time/op delta
CodeEncoder-8 2.48ms ± 9% 2.07ms ± 4% -16.36% (p=0.008 n=5+5)
CodeMarshal-8 2.58ms ± 1% 2.25ms ± 3% -12.88% (p=0.008 n=5+5)
MarshalBytes/32-8 265ns ± 2% 255ns ± 1% -3.70% (p=0.008 n=5+5)
MarshalBytes/256-8 637ns ± 2% 624ns ± 2% ~ (p=0.071 n=5+5)
MarshalBytes/4096-8 6.23µs ±12% 5.64µs ± 1% -9.47% (p=0.008 n=5+5)
CodeDecoder-8 11.0ms ± 3% 10.2ms ± 1% -7.21% (p=0.008 n=5+5)
UnicodeDecoder-8 259ns ± 0% 247ns ± 1% -4.63% (p=0.008 n=5+5)
DecoderStream-8 210ns ± 1% 206ns ± 1% -1.81% (p=0.016 n=5+5)
CodeUnmarshal-8 11.4ms ± 2% 11.2ms ± 4% ~ (p=0.222 n=5+5)
CodeUnmarshalReuse-8 10.3ms ± 1% 10.1ms ± 1% -2.30% (p=0.016 n=4+5)
UnmarshalString-8 120ns ± 2% 122ns ±10% ~ (p=0.913 n=5+5)
UnmarshalFloat64-8 112ns ± 8% 110ns ± 5% ~ (p=0.452 n=5+5)
UnmarshalInt64-8 90.5ns ± 2% 90.3ns ± 1% ~ (p=1.000 n=5+5)
Issue10335-8 159ns ± 3% 156ns ± 2% ~ (p=0.119 n=5+5)
Unmapped-8 357ns ± 5% 365ns ± 6% ~ (p=0.762 n=5+5)
TypeFieldsCache/MissTypes1-8 9.55µs ± 1% 9.27µs ± 7% ~ (p=0.151 n=5+5)
TypeFieldsCache/MissTypes10-8 35.6µs ± 8% 36.3µs ± 9% ~ (p=0.690 n=5+5)
TypeFieldsCache/MissTypes100-8 190µs ± 3% 184µs ± 9% ~ (p=0.310 n=5+5)
TypeFieldsCache/MissTypes1000-8 1.67ms ± 3% 1.58ms ± 6% -5.78% (p=0.032 n=5+5)
TypeFieldsCache/MissTypes10000-8 15.8ms ± 2% 17.2ms ± 9% +8.71% (p=0.008 n=5+5)
TypeFieldsCache/MissTypes100000-8 167ms ± 4% 162ms ± 7% ~ (p=0.548 n=5+5)
TypeFieldsCache/MissTypes1000000-8 1.69s ±26% 1.63s ±27% ~ (p=0.421 n=5+5)
TypeFieldsCache/HitTypes1-8 10.1ns ± 6% 8.8ns ± 2% -12.84% (p=0.008 n=5+5)
TypeFieldsCache/HitTypes10-8 9.10ns ± 2% 8.49ns ± 3% -6.72% (p=0.008 n=5+5)
TypeFieldsCache/HitTypes100-8 9.51ns ± 4% 9.26ns ± 2% ~ (p=0.095 n=5+5)
TypeFieldsCache/HitTypes1000-8 9.06ns ± 6% 8.96ns ± 3% ~ (p=1.000 n=5+5)
TypeFieldsCache/HitTypes10000-8 9.02ns ± 4% 8.36ns ± 6% -7.29% (p=0.032 n=5+5)
TypeFieldsCache/HitTypes100000-8 9.29ns ± 4% 9.52ns ± 6% ~ (p=0.151 n=5+5)
TypeFieldsCache/HitTypes1000000-8 10.7ns ± 7% 9.5ns ± 9% ~ (p=0.095 n=5+5)
NumberIsValid-8 17.3ns ± 6% 16.8ns ± 5% ~ (p=0.095 n=5+5)
NumberIsValidRegexp-8 384ns ± 3% 379ns ± 1% ~ (p=0.452 n=5+5)
EncoderEncode-8 70.4ns ± 2% 68.1ns ± 3% ~ (p=0.056 n=5+5)

name old speed new speed delta
CodeEncoder-8 784MB/s ± 9% 936MB/s ± 4% +19.28% (p=0.008 n=5+5)
CodeMarshal-8 752MB/s ± 1% 863MB/s ± 3% +14.82% (p=0.008 n=5+5)
CodeDecoder-8 177MB/s ± 3% 191MB/s ± 1% +7.71% (p=0.008 n=5+5)
UnicodeDecoder-8 54.0MB/s ± 1% 56.6MB/s ± 1% +4.81% (p=0.008 n=5+5)
CodeUnmarshal-8 170MB/s ± 2% 173MB/s ± 4% ~ (p=0.222 n=5+5)
CodeUnmarshalReuse-8 188MB/s ± 1% 192MB/s ± 1% +2.35% (p=0.016 n=4+5)

name old alloc/op new alloc/op delta
CodeEncoder-8 5.80B ±66% 4.60B ±57% ~ (p=0.238 n=5+5)
CodeMarshal-8 1.94MB ± 0% 1.96MB ± 3% ~ (p=0.190 n=4+5)
CodeDecoder-8 2.71MB ± 3% 2.64MB ± 1% -2.42% (p=0.008 n=5+5)
UnicodeDecoder-8 36.0B ± 0% 36.0B ± 0% ~ (all equal)
DecoderStream-8 8.00B ± 0% 8.00B ± 0% ~ (all equal)
CodeUnmarshal-8 3.28MB ± 0% 3.28MB ± 0% +0.00% (p=0.032 n=5+5)
CodeUnmarshalReuse-8 1.93MB ± 0% 1.93MB ± 0% ~ (p=0.056 n=5+5)
UnmarshalString-8 192B ± 0% 192B ± 0% ~ (all equal)
UnmarshalFloat64-8 180B ± 0% 180B ± 0% ~ (all equal)
UnmarshalInt64-8 176B ± 0% 176B ± 0% ~ (all equal)
Issue10335-8 200B ± 0% 200B ± 0% ~ (all equal)
Unmapped-8 232B ± 0% 232B ± 0% ~ (all equal)
EncoderEncode-8 0.00B 0.00B ~ (all equal)

name old allocs/op new allocs/op delta
CodeEncoder-8 0.00 0.00 ~ (all equal)
CodeMarshal-8 1.00 ± 0% 1.00 ± 0% ~ (all equal)
CodeDecoder-8 77.5k ± 0% 77.4k ± 0% -0.12% (p=0.008 n=5+5)
UnicodeDecoder-8 2.00 ± 0% 2.00 ± 0% ~ (all equal)
DecoderStream-8 1.00 ± 0% 1.00 ± 0% ~ (all equal)
CodeUnmarshal-8 92.7k ± 0% 92.7k ± 0% ~ (all equal)
CodeUnmarshalReuse-8 77.4k ± 0% 77.4k ± 0% ~ (p=0.063 n=5+5)
UnmarshalString-8 2.00 ± 0% 2.00 ± 0% ~ (all equal)
UnmarshalFloat64-8 2.00 ± 0% 2.00 ± 0% ~ (all equal)
UnmarshalInt64-8 1.00 ± 0% 1.00 ± 0% ~ (all equal)
Issue10335-8 3.00 ± 0% 3.00 ± 0% ~ (all equal)
Unmapped-8 4.00 ± 0% 4.00 ± 0% ~ (all equal)
EncoderEncode-8 0.00 0.00 ~ (all equal)

---

Please don’t reply on this GitHub thread. Visit golang.org/cl/194338.
After addressing review feedback, remember to publish your drafts!

[gopherbot]

Message from Daniel Martí:

Patch Set 1:

Do you think this is better than CL 183228? If not, let's focus on that CL instead.

---

Please don’t reply on this GitHub thread. Visit golang.org/cl/194338.
After addressing review feedback, remember to publish your drafts!

[gopherbot]

Message from Andreas Goetz:

Patch Set 1:

Patch Set 1:

Do you think this is better than CL 183228? If not, let's focus on that CL instead.

It's a three line problem, so not much possible difference. I think this CL is "better" in that its puts the reset logic in a single place while in CL 183228 its scattered around pool consumers. This one has 1 line less code.

Whatever CL implements this change the form of this one is slightly cleaner imho.

---

Please don’t reply on this GitHub thread. Visit golang.org/cl/194338.
After addressing review feedback, remember to publish your drafts!

[gopherbot]

Message from Daniel Martí:

Patch Set 1: Code-Review+2

I agree this is slightly better. And like Brad says, more consistent. LGTM

---

Please don’t reply on this GitHub thread. Visit golang.org/cl/194338.
After addressing review feedback, remember to publish your drafts!

[gopherbot]

This PR is being closed because golang.org/cl/194338 has been merged.