perf(http3): eliminate allocations with zero-copy Bytes#3010
Merged
Conversation
Fixes issue mozilla#2854 by implementing a zero-copy alternative to the closed PR mozilla#2988. Instead of using the `Bytes` crate (which has `Arc` overhead), this introduces a lightweight `DatagramPayload` wrapper that stores the original `Vec<u8>` with a payload offset, avoiding both memory allocations and memory copies. Key improvements: - WebTransport datagrams: 0 allocations, 0 copies (was 1 allocation via `to_vec()`) - Connect UDP datagrams: 0 allocations, 0 copies (was 1 allocation via `to_vec()`) - No external dependencies or `Arc` overhead - Efficient slice access via offset calculation - Maintains API compatibility through trait implementations CC @mxinden
Contributor
There was a problem hiding this comment.
Pull Request Overview
This PR implements a zero-copy DatagramPayload wrapper to eliminate memory allocations when handling HTTP/3 datagrams for WebTransport and Connect UDP sessions. The change replaces Vec<u8> with a custom DatagramPayload struct that stores the original data buffer with a payload offset, avoiding the need for to_vec() calls.
Key changes:
- Introduces
DatagramPayloadstruct with offset-based payload access - Updates all datagram-related APIs to use
DatagramPayloadinstead ofVec<u8> - Modifies datagram handling to pass ownership of the original buffer with offset information
Reviewed Changes
Copilot reviewed 10 out of 10 changed files in this pull request and generated 3 comments.
Show a summary per file
| File | Description |
|---|---|
| neqo-http3/src/lib.rs | Adds the new DatagramPayload struct with comprehensive trait implementations |
| neqo-http3/src/connection.rs | Updates datagram handling to pass owned Vec with offset instead of copying |
| neqo-http3/src/features/extended_connect/session.rs | Replaces datagram() with datagram_owned() to handle owned data with offset |
| neqo-http3/src/server_events.rs | Updates event structs and methods to use DatagramPayload |
| neqo-http3/src/client_events.rs | Updates client event structs and methods to use DatagramPayload |
| neqo-http3/src/server_connection_events.rs | Updates server connection events to use DatagramPayload |
| neqo-http3/src/features/extended_connect/mod.rs | Updates trait signature to use DatagramPayload |
| neqo-http3/src/connection_server.rs | Removes reference when passing datagram to handler |
| neqo-http3/src/connection_client.rs | Removes reference when passing datagram to handler |
| neqo-http3/src/features/extended_connect/tests/webtransport/mod.rs | Updates test assertions to use as_ref() for comparison |
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
Codecov Report❌ Patch coverage is Additional details and impacted files@@ Coverage Diff @@
## main #3010 +/- ##
==========================================
- Coverage 93.37% 93.35% -0.03%
==========================================
Files 123 124 +1
Lines 35887 35967 +80
Branches 35887 35967 +80
==========================================
+ Hits 33511 33578 +67
- Misses 1533 1546 +13
Partials 843 843
|
larseggert
changed the title
DatagramPayload`DatagramPayload
Contributor
|
Bencher Report| Branch | fix-2854-v2 |
| Testbed | On-prem |
🚨 1 Alert
| Benchmark | Measure Units | View | Benchmark Result (Result Δ%) | Upper Boundary (Limit %) |
|---|---|---|---|---|
| decode 4096 bytes, mask 3f | Latency microseconds (µs) | 📈 plot 🚷 threshold 🚨 alert (🔔) | 8.38 µs(+1.05%)Baseline: 8.30 µs | 8.35 µs (100.44%) |
Click to view all benchmark results
| Benchmark | Latency | Benchmark Result nanoseconds (ns) (Result Δ%) | Upper Boundary nanoseconds (ns) (Limit %) |
|---|---|---|---|
| 1-conn/1-100mb-req/mtu-1504 (aka. Upload)/client | 📈 view plot 🚷 view threshold | 199,620,000.00 ns(-4.49%)Baseline: 209,009,375.00 ns | 218,003,224.46 ns (91.57%) |
| 1-conn/1-100mb-resp/mtu-1504 (aka. Download)/client | 📈 view plot 🚷 view threshold | 193,950,000.00 ns(-4.47%)Baseline: 203,033,465.91 ns | 212,898,570.27 ns (91.10%) |
| 1-conn/1-1b-resp/mtu-1504 (aka. HPS)/client | 📈 view plot 🚷 view threshold | 28,423,000.00 ns(+0.04%)Baseline: 28,412,869.32 ns | 28,869,091.75 ns (98.45%) |
| 1-conn/10_000-parallel-1b-resp/mtu-1504 (aka. RPS)/client | 📈 view plot 🚷 view threshold | 287,430,000.00 ns(-2.50%)Baseline: 294,794,772.73 ns | 306,070,236.30 ns (93.91%) |
| 1-streams/each-1000-bytes/simulated-time | 📈 view plot 🚷 view threshold | 119,040,000.00 ns(+0.62%)Baseline: 118,310,056.82 ns | 120,887,916.37 ns (98.47%) |
| 1-streams/each-1000-bytes/wallclock-time | 📈 view plot 🚷 view threshold | 586,670.00 ns(-1.96%)Baseline: 598,389.38 ns | 623,192.47 ns (94.14%) |
| 1000-streams/each-1-bytes/simulated-time | 📈 view plot 🚷 view threshold | 14,994,000,000.00 ns(+0.01%)Baseline: 14,991,823,863.64 ns | 15,010,334,126.88 ns (99.89%) |
| 1000-streams/each-1-bytes/wallclock-time | 📈 view plot 🚷 view threshold | 13,663,000.00 ns(-3.91%)Baseline: 14,219,079.55 ns | 14,990,431.89 ns (91.14%) |
| 1000-streams/each-1000-bytes/simulated-time | 📈 view plot 🚷 view threshold | 19,049,000,000.00 ns(+0.74%)Baseline: 18,909,653,409.09 ns | 19,156,534,875.56 ns (99.44%) |
| 1000-streams/each-1000-bytes/wallclock-time | 📈 view plot 🚷 view threshold | 48,923,000.00 ns(-6.49%)Baseline: 52,316,011.36 ns | 58,757,035.26 ns (83.26%) |
| RxStreamOrderer::inbound_frame() | 📈 view plot 🚷 view threshold | 108,490,000.00 ns(-1.23%)Baseline: 109,846,363.64 ns | 112,004,019.57 ns (96.86%) |
| coalesce_acked_from_zero 1+1 entries | 📈 view plot 🚷 view threshold | 88.56 ns(-0.08%)Baseline: 88.63 ns | 89.31 ns (99.17%) |
| coalesce_acked_from_zero 10+1 entries | 📈 view plot 🚷 view threshold | 105.56 ns(-0.51%)Baseline: 106.10 ns | 107.09 ns (98.57%) |
| coalesce_acked_from_zero 1000+1 entries | 📈 view plot 🚷 view threshold | 88.87 ns(-1.07%)Baseline: 89.84 ns | 94.49 ns (94.06%) |
| coalesce_acked_from_zero 3+1 entries | 📈 view plot 🚷 view threshold | 106.14 ns(-0.45%)Baseline: 106.62 ns | 107.58 ns (98.67%) |
| decode 1048576 bytes, mask 3f | 📈 view plot 🚷 view threshold | 1,593,300.00 ns(+0.05%)Baseline: 1,592,547.73 ns | 1,599,591.84 ns (99.61%) |
| decode 1048576 bytes, mask 7f | 📈 view plot 🚷 view threshold | 5,053,300.00 ns(-0.08%)Baseline: 5,057,278.98 ns | 5,077,128.25 ns (99.53%) |
| decode 1048576 bytes, mask ff | 📈 view plot 🚷 view threshold | 3,036,500.00 ns(+0.16%)Baseline: 3,031,650.57 ns | 3,043,537.16 ns (99.77%) |
| decode 4096 bytes, mask 3f | 📈 view plot 🚷 view threshold 🚨 view alert (🔔) | 8,383.90 ns(+1.05%)Baseline: 8,297.09 ns | 8,347.38 ns (100.44%) |
| decode 4096 bytes, mask 7f | 📈 view plot 🚷 view threshold | 19,994.00 ns(-0.07%)Baseline: 20,007.75 ns | 20,086.37 ns (99.54%) |
| decode 4096 bytes, mask ff | 📈 view plot 🚷 view threshold | 11,636.00 ns(-0.82%)Baseline: 11,732.65 ns | 11,977.40 ns (97.15%) |
| sent::Packets::take_ranges | 📈 view plot 🚷 view threshold | 4,571.20 ns(-3.70%)Baseline: 4,746.71 ns | 4,990.23 ns (91.60%) |
| transfer/pacing-false/same-seed/simulated-time/run | 📈 view plot 🚷 view threshold | 25,710,000,000.00 ns(+1.84%)Baseline: 25,245,000,000.00 ns | 25,733,296,634.69 ns (99.91%) |
| transfer/pacing-false/same-seed/wallclock-time/run | 📈 view plot 🚷 view threshold | 25,638,000.00 ns(-1.50%)Baseline: 26,028,500.00 ns | 27,081,526.02 ns (94.67%) |
| transfer/pacing-false/varying-seeds/simulated-time/run | 📈 view plot 🚷 view threshold | 25,185,000,000.00 ns(+0.07%)Baseline: 25,166,844,827.59 ns | 25,211,663,017.04 ns (99.89%) |
| transfer/pacing-false/varying-seeds/wallclock-time/run | 📈 view plot 🚷 view threshold | 25,049,000.00 ns(-4.64%)Baseline: 26,268,856.32 ns | 27,610,125.20 ns (90.72%) |
| transfer/pacing-true/same-seed/simulated-time/run | 📈 view plot 🚷 view threshold | 25,675,000,000.00 ns(+0.28%)Baseline: 25,602,500,000.00 ns | 25,678,632,271.00 ns (99.99%) |
| transfer/pacing-true/same-seed/wallclock-time/run | 📈 view plot 🚷 view threshold | 26,570,000.00 ns(-3.24%)Baseline: 27,460,000.00 ns | 28,771,349.63 ns (92.35%) |
| transfer/pacing-true/varying-seeds/simulated-time/run | 📈 view plot 🚷 view threshold | 25,009,000,000.00 ns(+0.06%)Baseline: 24,994,063,218.39 ns | 25,043,698,696.08 ns (99.86%) |
| transfer/pacing-true/varying-seeds/wallclock-time/run | 📈 view plot 🚷 view threshold | 25,554,000.00 ns(-4.65%)Baseline: 26,799,793.10 ns | 28,176,607.47 ns (90.69%) |
Co-authored-by: Copilot <[email protected]> Signed-off-by: Lars Eggert <[email protected]>
Co-authored-by: Copilot <[email protected]> Signed-off-by: Lars Eggert <[email protected]>
Contributor
There was a problem hiding this comment.
Pull Request Overview
Copilot reviewed 10 out of 10 changed files in this pull request and generated 2 comments.
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
Contributor
|
| Branch | fix-2854-v2 |
| Testbed | On-prem |
Click to view all benchmark results
| Benchmark | Latency | Benchmark Result milliseconds (ms) (Result Δ%) | Upper Boundary milliseconds (ms) (Limit %) |
|---|---|---|---|
| google vs. neqo (cubic, paced) | 📈 view plot 🚷 view threshold | 278.80 ms(+0.40%)Baseline: 277.69 ms | 280.28 ms (99.47%) |
| Benchmark | Latency | Benchmark Result milliseconds (ms) (Result Δ%) | Upper Boundary milliseconds (ms) (Limit %) |
|---|---|---|---|
| msquic vs. neqo (cubic, paced) | 📈 view plot 🚷 view threshold | 182.69 ms(-5.60%)Baseline: 193.53 ms | 226.60 ms (80.62%) |
| Benchmark | Latency | Benchmark Result milliseconds (ms) (Result Δ%) | Upper Boundary milliseconds (ms) (Limit %) |
|---|---|---|---|
| neqo vs. google (cubic, paced) | 📈 view plot 🚷 view threshold | 751.91 ms(-0.75%)Baseline: 757.57 ms | 764.92 ms (98.30%) |
| Benchmark | Latency | Benchmark Result milliseconds (ms) (Result Δ%) | Upper Boundary milliseconds (ms) (Limit %) |
|---|---|---|---|
| neqo vs. msquic (cubic, paced) | 📈 view plot 🚷 view threshold | 159.17 ms(+1.26%)Baseline: 157.20 ms | 159.51 ms (99.79%) |
| Benchmark | Latency | Benchmark Result milliseconds (ms) (Result Δ%) | Upper Boundary milliseconds (ms) (Limit %) |
|---|---|---|---|
| neqo vs. neqo (cubic) | 📈 view plot 🚷 view threshold | 90.53 ms(-0.17%)Baseline: 90.69 ms | 94.49 ms (95.81%) |
| Benchmark | Latency | Benchmark Result milliseconds (ms) (Result Δ%) | Upper Boundary milliseconds (ms) (Limit %) |
|---|---|---|---|
| neqo vs. neqo (cubic, paced) | 📈 view plot 🚷 view threshold | 92.62 ms(+0.78%)Baseline: 91.91 ms | 95.61 ms (96.88%) |
| Benchmark | Latency | Benchmark Result milliseconds (ms) (Result Δ%) | Upper Boundary milliseconds (ms) (Limit %) |
|---|---|---|---|
| neqo vs. neqo (reno) | 📈 view plot 🚷 view threshold | 89.38 ms(-1.36%)Baseline: 90.61 ms | 94.00 ms (95.08%) |
| Benchmark | Latency | Benchmark Result milliseconds (ms) (Result Δ%) | Upper Boundary milliseconds (ms) (Limit %) |
|---|---|---|---|
| neqo vs. neqo (reno, paced) | 📈 view plot 🚷 view threshold | 91.54 ms(-0.33%)Baseline: 91.84 ms | 95.51 ms (95.84%) |
| Benchmark | Latency | Benchmark Result milliseconds (ms) (Result Δ%) | Upper Boundary milliseconds (ms) (Limit %) |
|---|---|---|---|
| neqo vs. quiche (cubic, paced) | 📈 view plot 🚷 view threshold | 193.64 ms(-0.06%)Baseline: 193.76 ms | 197.21 ms (98.19%) |
| Benchmark | Latency | Benchmark Result milliseconds (ms) (Result Δ%) | Upper Boundary milliseconds (ms) (Limit %) |
|---|---|---|---|
| neqo vs. s2n (cubic, paced) | 📈 view plot 🚷 view threshold | 219.63 ms(-0.53%)Baseline: 220.81 ms | 223.63 ms (98.21%) |
| Benchmark | Latency | Benchmark Result milliseconds (ms) (Result Δ%) | Upper Boundary milliseconds (ms) (Limit %) |
|---|---|---|---|
| quiche vs. neqo (cubic, paced) | 📈 view plot 🚷 view threshold | 152.44 ms(-0.06%)Baseline: 152.54 ms | 158.23 ms (96.34%) |
| Benchmark | Latency | Benchmark Result milliseconds (ms) (Result Δ%) | Upper Boundary milliseconds (ms) (Limit %) |
|---|---|---|---|
| s2n vs. neqo (cubic, paced) | 📈 view plot 🚷 view threshold | 174.02 ms(+0.05%)Baseline: 173.94 ms | 178.09 ms (97.72%) |
larseggert
requested review from
KershawChang,
martinthomson and
mxinden
as code owners
Contributor
There was a problem hiding this comment.
Pull Request Overview
Copilot reviewed 10 out of 10 changed files in this pull request and generated 2 comments.
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
Co-authored-by: Copilot <[email protected]> Signed-off-by: Lars Eggert <[email protected]>
Contributor
There was a problem hiding this comment.
Pull Request Overview
Copilot reviewed 10 out of 10 changed files in this pull request and generated no new comments.
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
martinthomson
approved these changes
Sep 25, 2025
Member
martinthomson
left a comment
martinthomson
left a comment
There was a problem hiding this comment.
I think that I found a bug, not exactly in your code, but highlighted by it. Still, this seems like a good change.
larseggert
commented
Sep 26, 2025
mxinden
reviewed
Sep 29, 2025
Contributor
There was a problem hiding this comment.
Pull Request Overview
Copilot reviewed 14 out of 15 changed files in this pull request and generated 2 comments.
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
Contributor
There was a problem hiding this comment.
Pull Request Overview
Copilot reviewed 15 out of 15 changed files in this pull request and generated 2 comments.
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
mxinden
approved these changes
Oct 9, 2025
Member
mxinden
left a comment
mxinden
left a comment
There was a problem hiding this comment.
Other than these minor comments, this looks good to me.
mxinden
approved these changes
Oct 10, 2025
Contributor
There was a problem hiding this comment.
Pull Request Overview
Copilot reviewed 15 out of 15 changed files in this pull request and generated 2 comments.
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
larseggert
commented
Oct 13, 2025
Co-authored-by: Copilot <[email protected]> Signed-off-by: Lars Eggert <[email protected]>
Contributor
There was a problem hiding this comment.
Pull Request Overview
Copilot reviewed 15 out of 15 changed files in this pull request and generated 2 comments.
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
Contributor
There was a problem hiding this comment.
Pull Request Overview
Copilot reviewed 15 out of 15 changed files in this pull request and generated 1 comment.
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
Contributor
Failed Interop TestsQUIC Interop Runner, client vs. server neqo-latest as client
neqo-latest as server
All resultsSucceeded Interop TestsQUIC Interop Runner, client vs. server neqo-latest as client
neqo-latest as server
Unsupported Interop TestsQUIC Interop Runner, client vs. server neqo-latest as client
neqo-latest as server
|
Contributor
Client/server transfer resultsPerformance differences relative to 3d2948d. Transfer of 33554432 bytes over loopback, min. 100 runs. All unit-less numbers are in milliseconds.
Download data for |
Contributor
Benchmark resultsPerformance differences relative to 3d2948d. 1-conn/1-100mb-resp/mtu-1504 (aka. Download)/client: Change within noise threshold. time: [193.67 ms 193.95 ms 194.23 ms]
thrpt: [514.84 MiB/s 515.60 MiB/s 516.35 MiB/s]
change:
time: [−0.8738% −0.6331% −0.4200%] (p = 0.00 < 0.05)
thrpt: [+0.4218% +0.6372% +0.8815%]
1-conn/10_000-parallel-1b-resp/mtu-1504 (aka. RPS)/client: No change in performance detected. time: [285.77 ms 287.43 ms 289.13 ms]
thrpt: [34.587 Kelem/s 34.791 Kelem/s 34.993 Kelem/s]
change:
time: [−1.2586% −0.4841% +0.3170%] (p = 0.23 > 0.05)
thrpt: [−0.3160% +0.4864% +1.2746%]
1-conn/1-1b-resp/mtu-1504 (aka. HPS)/client: No change in performance detected. time: [28.352 ms 28.423 ms 28.501 ms]
thrpt: [35.086 B/s 35.183 B/s 35.271 B/s]
change:
time: [−0.4448% +0.0369% +0.5069%] (p = 0.88 > 0.05)
thrpt: [−0.5043% −0.0369% +0.4468%]
1-conn/1-100mb-req/mtu-1504 (aka. Upload)/client: No change in performance detected. time: [199.27 ms 199.62 ms 200.05 ms]
thrpt: [499.88 MiB/s 500.95 MiB/s 501.82 MiB/s]
change:
time: [−0.2852% −0.0198% +0.2459%] (p = 0.89 > 0.05)
thrpt: [−0.2453% +0.0198% +0.2860%]
decode 4096 bytes, mask ff: No change in performance detected. time: [11.597 µs 11.636 µs 11.679 µs]
change: [−0.5655% −0.1027% +0.3064%] (p = 0.66 > 0.05)
decode 1048576 bytes, mask ff: No change in performance detected. time: [3.0231 ms 3.0365 ms 3.0535 ms]
change: [−0.4249% +0.1564% +0.7337%] (p = 0.62 > 0.05)
decode 4096 bytes, mask 7f: No change in performance detected. time: [19.943 µs 19.994 µs 20.050 µs]
change: [−0.7023% −0.1380% +0.4892%] (p = 0.66 > 0.05)
decode 1048576 bytes, mask 7f: No change in performance detected. time: [5.0423 ms 5.0533 ms 5.0650 ms]
change: [−0.4083% −0.0736% +0.2237%] (p = 0.67 > 0.05)
decode 4096 bytes, mask 3f: No change in performance detected. time: [8.2735 µs 8.3839 µs 8.5931 µs]
change: [−3.2234% −0.4227% +2.2355%] (p = 0.80 > 0.05)
decode 1048576 bytes, mask 3f: No change in performance detected. time: [1.5865 ms 1.5933 ms 1.6004 ms]
change: [−0.2760% +0.2505% +0.7841%] (p = 0.38 > 0.05)
1-streams/each-1000-bytes/wallclock-time: No change in performance detected. time: [584.00 µs 586.67 µs 589.62 µs]
change: [−0.7850% −0.2494% +0.3293%] (p = 0.40 > 0.05)
1000-streams/each-1-bytes/wallclock-time: No change in performance detected. time: [13.616 ms 13.663 ms 13.721 ms]
change: [−0.0720% +0.3384% +0.7826%] (p = 0.13 > 0.05)
1000-streams/each-1000-bytes/wallclock-time: No change in performance detected. time: [48.752 ms 48.923 ms 49.094 ms]
change: [−0.3027% +0.1962% +0.7040%] (p = 0.45 > 0.05)
coalesce_acked_from_zero 1+1 entries: No change in performance detected. time: [88.050 ns 88.563 ns 89.270 ns]
change: [−0.3826% +0.2811% +1.0626%] (p = 0.48 > 0.05)
coalesce_acked_from_zero 3+1 entries: No change in performance detected. time: [105.82 ns 106.14 ns 106.49 ns]
change: [−0.6431% −0.1753% +0.2647%] (p = 0.46 > 0.05)
coalesce_acked_from_zero 10+1 entries: No change in performance detected. time: [105.21 ns 105.56 ns 106.00 ns]
change: [−0.4737% −0.0516% +0.4260%] (p = 0.82 > 0.05)
coalesce_acked_from_zero 1000+1 entries: No change in performance detected. time: [88.752 ns 88.872 ns 89.015 ns]
change: [−0.6650% +0.1796% +1.1427%] (p = 0.71 > 0.05)
RxStreamOrderer::inbound_frame(): Change within noise threshold. time: [108.42 ms 108.49 ms 108.56 ms]
change: [−1.0062% −0.7124% −0.5013%] (p = 0.00 < 0.05)
sent::Packets::take_ranges: No change in performance detected. time: [4.4988 µs 4.5712 µs 4.6306 µs]
change: [−5.6252% −2.2346% +1.0580%] (p = 0.20 > 0.05)
transfer/pacing-false/varying-seeds/wallclock-time/run: No change in performance detected. time: [25.005 ms 25.049 ms 25.104 ms]
change: [−0.1686% +0.0520% +0.2935%] (p = 0.69 > 0.05)
transfer/pacing-false/varying-seeds/simulated-time/run: No change in performance detected. time: [25.154 s 25.185 s 25.216 s]
thrpt: [162.44 KiB/s 162.64 KiB/s 162.84 KiB/s]
change:
time: [−0.0799% +0.1115% +0.3030%] (p = 0.25 > 0.05)
thrpt: [−0.3021% −0.1114% +0.0800%]
transfer/pacing-true/varying-seeds/wallclock-time/run: No change in performance detected. time: [25.491 ms 25.554 ms 25.618 ms]
change: [−0.2540% +0.1091% +0.4740%] (p = 0.56 > 0.05)
transfer/pacing-true/varying-seeds/simulated-time/run: No change in performance detected. time: [24.972 s 25.009 s 25.047 s]
thrpt: [163.53 KiB/s 163.78 KiB/s 164.03 KiB/s]
change:
time: [−0.1066% +0.0973% +0.2930%] (p = 0.36 > 0.05)
thrpt: [−0.2921% −0.0972% +0.1067%]
transfer/pacing-false/same-seed/wallclock-time/run: Change within noise threshold. time: [25.618 ms 25.638 ms 25.658 ms]
change: [+0.6091% +0.8101% +0.9669%] (p = 0.00 < 0.05)
transfer/pacing-false/same-seed/simulated-time/run: No change in performance detected. time: [25.710 s 25.710 s 25.710 s]
thrpt: [159.31 KiB/s 159.31 KiB/s 159.31 KiB/s]
change:
time: [+0.0000% +0.0000% +0.0000%] (p = NaN > 0.05)
thrpt: [+0.0000% +0.0000% +0.0000%]
transfer/pacing-true/same-seed/wallclock-time/run: Change within noise threshold. time: [26.545 ms 26.570 ms 26.600 ms]
change: [+0.4713% +0.6253% +0.7668%] (p = 0.00 < 0.05)
transfer/pacing-true/same-seed/simulated-time/run: No change in performance detected. time: [25.675 s 25.675 s 25.675 s]
thrpt: [159.53 KiB/s 159.53 KiB/s 159.53 KiB/s]
change:
time: [+0.0000% +0.0000% +0.0000%] (p = NaN > 0.05)
thrpt: [+0.0000% +0.0000% +0.0000%]
Download data for |
Merged
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Fixes issue #2854 by implementing a zero-copy alternative to the closed PR #2988.
Alternative to #2988
CC @mxinden