fix(engine): render() can only return qualifying templates

[caridy]

Details

Previously, we supported render() to return undefined or qualifying template. With this PR, we will enforce that if there is a render method, it should return a qualifying template.

Does this PR introduce a breaking change?

• Yes

[salesforce-best-lwc-internal]

Benchmark results

Base commit: 00a13cc | Target commit: 25013b0

lwc-engine-benchmark

table-append-1k	metric	base(00a13cc)	target(25013b0)	trend
benchmark-table/append/1k	duration	156.25 (±4.80 ms)	150.45 (±5.20 ms)	-5.8ms (3.7%) 👍

table-clear-1k	metric	base(00a13cc)	target(25013b0)	trend
benchmark-table/clear/1k	duration	6.60 (±0.30 ms)	6.50 (±0.40 ms)	-0.1ms (1.5%) 👌

table-create-10k	metric	base(00a13cc)	target(25013b0)	trend
benchmark-table/create/10k	duration	879.85 (±5.55 ms)	877.30 (±7.25 ms)	-2.5ms (0.3%) 👌

table-create-1k	metric	base(00a13cc)	target(25013b0)	trend
benchmark-table/create/1k	duration	114.90 (±2.40 ms)	115.75 (±2.70 ms)	+0.8ms (0.7%) 👌

table-update-10th-1k	metric	base(00a13cc)	target(25013b0)	trend
benchmark-table/update-10th/1k	duration	86.05 (±3.00 ms)	79.85 (±2.60 ms)	-6.2ms (7.2%) 👌

tablecmp-append-1k	metric	base(00a13cc)	target(25013b0)	trend
benchmark-table-component/append/1k	duration	215.90 (±13.20 ms)	240.10 (±13.80 ms)	+24.2ms (11.2%) 👎

tablecmp-clear-1k	metric	base(00a13cc)	target(25013b0)	trend
benchmark-table-component/clear/1k	duration	11.40 (±1.30 ms)	11.80 (±1.50 ms)	+0.4ms (3.5%) 👌

tablecmp-create-10k	metric	base(00a13cc)	target(25013b0)	trend
benchmark-table-component/create/10k	duration	1656.90 (±12.90 ms)	1662.50 (±23.85 ms)	+5.6ms (0.3%) 👎

tablecmp-create-1k	metric	base(00a13cc)	target(25013b0)	trend
benchmark-table-component/create/1k	duration	198.90 (±4.80 ms)	200.60 (±4.10 ms)	+1.7ms (0.9%) 👌

tablecmp-update-10th-1k	metric	base(00a13cc)	target(25013b0)	trend
benchmark-table-component/update-10th/1k	duration	69.35 (±4.25 ms)	72.95 (±5.15 ms)	+3.6ms (5.2%) 👌

wc-append-1k	metric	base(00a13cc)	target(25013b0)	trend
benchmark-table-wc/append/1k	duration	231.25 (±9.40 ms)	246.40 (±14.75 ms)	+15.1ms (6.6%) 👎

wc-clear-1k	metric	base(00a13cc)	target(25013b0)	trend
benchmark-table-wc/clear/1k	duration	22.15 (±2.55 ms)	23.60 (±2.10 ms)	+1.5ms (6.5%) 👎

wc-create-10k	metric	base(00a13cc)	target(25013b0)	trend
benchmark-table-wc/create/10k	duration	1685.95 (±52.35 ms)	1942.15 (±63.10 ms)	+256.2ms (15.2%) 👎

wc-create-1k	metric	base(00a13cc)	target(25013b0)	trend
benchmark-table-wc/create/1k	duration	206.80 (±5.20 ms)	227.60 (±4.30 ms)	+20.8ms (10.1%) 👎

wc-update-10th-1k	metric	base(00a13cc)	target(25013b0)	trend
benchmark-table-wc/update-10th/1k	duration	75.05 (±6.25 ms)	78.55 (±5.95 ms)	+3.5ms (4.7%) 👎

[kevinv11n]

Why does renderedCallback need to return this?

[caridy]

uff, my bad!

[ekashida]

Removed

[diervo]

@caridy do another empty commit to double check the numbers are not as odd.

[ekashida]

Changes look good but we'll need to exercise caution when releasing this.

[salesforce-best-lwc-internal]

Benchmark results

Base commit: 00a13cc | Target commit: 4df6162

lwc-engine-benchmark

table-append-1k	metric	base(00a13cc)	target(4df6162)	trend
benchmark-table/append/1k	duration	156.25 (±4.80 ms)	146.95 (±4.45 ms)	-9.3ms (6.0%) 👍

table-clear-1k	metric	base(00a13cc)	target(4df6162)	trend
benchmark-table/clear/1k	duration	6.60 (±0.30 ms)	5.95 (±0.30 ms)	-0.6ms (9.8%) 👍

table-create-10k	metric	base(00a13cc)	target(4df6162)	trend
benchmark-table/create/10k	duration	879.85 (±5.55 ms)	879.65 (±6.30 ms)	-0.2ms (0.0%) 👌

table-create-1k	metric	base(00a13cc)	target(4df6162)	trend
benchmark-table/create/1k	duration	114.90 (±2.40 ms)	113.55 (±3.10 ms)	-1.4ms (1.2%) 👌

table-update-10th-1k	metric	base(00a13cc)	target(4df6162)	trend
benchmark-table/update-10th/1k	duration	86.05 (±3.00 ms)	75.75 (±1.50 ms)	-10.3ms (12.0%) 👍

tablecmp-append-1k	metric	base(00a13cc)	target(4df6162)	trend
benchmark-table-component/append/1k	duration	215.90 (±13.20 ms)	236.35 (±11.25 ms)	+20.4ms (9.5%) 👎

tablecmp-clear-1k	metric	base(00a13cc)	target(4df6162)	trend
benchmark-table-component/clear/1k	duration	11.40 (±1.30 ms)	11.80 (±1.40 ms)	+0.4ms (3.5%) 👌

tablecmp-create-10k	metric	base(00a13cc)	target(4df6162)	trend
benchmark-table-component/create/10k	duration	1656.90 (±12.90 ms)	1691.65 (±10.10 ms)	+34.8ms (2.1%) 👎

tablecmp-create-1k	metric	base(00a13cc)	target(4df6162)	trend
benchmark-table-component/create/1k	duration	198.90 (±4.80 ms)	205.35 (±4.75 ms)	+6.5ms (3.2%) 👎

tablecmp-update-10th-1k	metric	base(00a13cc)	target(4df6162)	trend
benchmark-table-component/update-10th/1k	duration	69.35 (±4.25 ms)	69.35 (±3.95 ms)	0.0ms (0.0%) 👌

wc-append-1k	metric	base(00a13cc)	target(4df6162)	trend
benchmark-table-wc/append/1k	duration	231.25 (±9.40 ms)	243.90 (±9.05 ms)	+12.7ms (5.5%) 👎

wc-clear-1k	metric	base(00a13cc)	target(4df6162)	trend
benchmark-table-wc/clear/1k	duration	22.15 (±2.55 ms)	21.90 (±2.40 ms)	-0.3ms (1.1%) 👌

wc-create-10k	metric	base(00a13cc)	target(4df6162)	trend
benchmark-table-wc/create/10k	duration	1685.95 (±52.35 ms)	1875.70 (±27.05 ms)	+189.8ms (11.3%) 👎

wc-create-1k	metric	base(00a13cc)	target(4df6162)	trend
benchmark-table-wc/create/1k	duration	206.80 (±5.20 ms)	216.35 (±3.05 ms)	+9.5ms (4.6%) 👎

wc-update-10th-1k	metric	base(00a13cc)	target(4df6162)	trend
benchmark-table-wc/update-10th/1k	duration	75.05 (±6.25 ms)	72.75 (±5.95 ms)	-2.3ms (3.1%) 👌

[caridy]

@diervo not sure how to justify that regression, this PR removes two conditions in prod mode. It should be faster :), and it is not about wc, but for any rendering process.

[ekashida]

@diervo We need to hold off on this one until after LGC finishes their long-awaited release. This change also needs to go in with 16867957.

[ekashida]

The necessary downstream changes have been made but we probably want to release this separately from the big CSS changes since it's always possible (though unlikely) that someone added a new render method that returns undefined.

[caridy]

ready to roll