Improve encoding of lists where first item can't fit in packet. #28346
Conversation
This file contains 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
pullapprove
bot
requested review from
andy31415,
andyg-apple,
anush-apple,
arkq,
carol-apple,
cecille,
chrisdecenzo,
chshu,
chulspro,
cliffamzn,
Damian-Nordic,
dhrishi,
electrocucaracha,
emargolis,
gjc13,
harsha-rajendran,
hawk248,
hicklin,
jepenven-silabs,
jmartinez-silabs,
jmeg-sfy,
joonhaengHeo,
jtung-apple,
kkasperczyk-no,
ksperling-apple,
lazarkov,
lpbeliveau-silabs and
LuDuda
pullapprove
bot
requested review from
saurabhst,
selissia,
sharadb-amazon,
tcarmelveilleux,
tecimovic,
tehampson,
tima-q,
tobiasgraf,
vivien-apple,
woody-apple,
younghak-hwang and
yufengwangca
|
PR #28346: Size comparison from 05e4cee to 351dc16 Increases (58 builds for bl602, bl702, bl702l, cc32xx, cyw30739, efr32, esp32, k32w, linux, mbed, nrfconnect, psoc6, qpg, telink)
Decreases (5 builds for bl602, bl702, bl702l, telink)
Full report (60 builds for bl602, bl702, bl702l, cc32xx, cyw30739, efr32, esp32, k32w, linux, mbed, nrfconnect, psoc6, qpg, telink)
|
tcarmelveilleux
approved these changes
Jul 28, 2023
yunhanw-google
approved these changes
Jul 28, 2023
When the first item of a list can't fit in the current packet, it's better to just send the packet without the list at all, then encode as many items as we can in the packet that follows before we start doing IB-per-item. Otherwise (before this change) we encode an empty list IB, then end up doing IB-per-item for the whole list.
bzbarsky-apple
force-pushed
the
no-output-empty-list-start
branch
from
351dc16
to
86fc2f6
Compare
|
PR #28346: Size comparison from 8363494 to 86fc2f6 Increases (59 builds for bl602, bl702, bl702l, cc32xx, cyw30739, efr32, esp32, k32w, linux, mbed, nrfconnect, psoc6, qpg, telink)
Decreases (3 builds for bl702, bl702l)
Full report (60 builds for bl602, bl702, bl702l, cc32xx, cyw30739, efr32, esp32, k32w, linux, mbed, nrfconnect, psoc6, qpg, telink)
|
bzbarsky-apple
force-pushed
the
no-output-empty-list-start
branch
from
86fc2f6
to
eafbb09
Compare
|
PR #28346: Size comparison from 8363494 to eafbb09 Increases (59 builds for bl602, bl702, bl702l, cc32xx, cyw30739, efr32, esp32, k32w, linux, mbed, nrfconnect, psoc6, qpg, telink)
Decreases (4 builds for bl702, bl702l, telink)
Full report (60 builds for bl602, bl702, bl702l, cc32xx, cyw30739, efr32, esp32, k32w, linux, mbed, nrfconnect, psoc6, qpg, telink)
|
jmartinez-silabs
approved these changes
Aug 7, 2023
abpoth
pushed a commit
to abpoth/connectedhomeip
that referenced
this pull request
Aug 15, 2023
…ect-chip#28346) * Improve encoding of lists where first item can't fit in packet. When the first item of a list can't fit in the current packet, it's better to just send the packet without the list at all, then encode as many items as we can in the packet that follows before we start doing IB-per-item. Otherwise (before this change) we encode an empty list IB, then end up doing IB-per-item for the whole list. * Improve unit tests.
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.
When the first item of a list can't fit in the current packet, it's better to just send the packet without the list at all, then encode as many items as we can in the packet that follows before we start doing IB-per-item.
Otherwise (before this change) we encode an empty list IB, then end up doing IB-per-item for the whole list.