Handle a lost standalone ack correctly.

[bzbarsky-apple]

If a standalone ack gets lost and then there is an application-level
reply to the message that triggered the standalone ack, we want to
piggyback an ack on that application-level reply as well. Otherwise
the other side can end up in a state where it has two outstanding
unacknowledged messages (the message we are replying to, and the reply
to our reply), which it can't handle.

The other change here is to fix ReliableMessageMgr to not crash if the
peer misbehaves and skips sending that piggyback ack described above.
Instead, we just error out from sending the response to the peer's
broken message.

Fixes #9796

Problem

See above.

Change overview

See above.

Testing

New unit test passes.

[todo]

A test that we should have but can't write with the existing

connectedhomeip/src/messaging/tests/TestReliableMessageProtocol.cpp

Lines 1369 to 1379 in b31a9c9

	* TODO: A test that we should have but can't write with the existing
	* infrastructure we have:
	*
	* 1. A sends message 1 to B
	* 2. B is slow to respond, A does a resend and the resend is delayed in the network.
	* 3. B responds with message 2, which acks message 1.
	* 4. A sends message 3 to B
	* 5. B sends standalone ack to message 3, which is lost
	* 6. The duplicate message from step 3 is delivered and triggers a standalone ack.
	* 7. B responds with message 4, which should carry a piggyback ack for message 3
	* (this is the part that needs testing!)

---

This comment was generated by todo based on a TODO comment in b31a9c9 in #9895. cc @bzbarsky-apple.

[github-actions]

Size increase report for "gn_qpg-example-build" from a62ee31

File	Section	File	VM
chip-qpg6100-lighting-example.out	.text	16	16

Full report output

BLOAT REPORT

Files found only in the build output:
    report.csv

Comparing ./master_artifact/chip-qpg6100-lighting-example.out.map and ./pull_artifact/chip-qpg6100-lighting-example.out.map:

BLOAT EXECUTION FAILED WITH CODE 1:
bloaty: unknown file type for file './pull_artifact/chip-qpg6100-lighting-example.out.map'

Comparing ./master_artifact/chip-qpg6100-lighting-example.out and ./pull_artifact/chip-qpg6100-lighting-example.out:

sections,vmsize,filesize
.debug_info,0,1837
.debug_loc,0,597
.debug_ranges,0,64
.debug_abbrev,0,55
.symtab,0,32
.debug_frame,0,16
.text,16,16
.debug_aranges,0,8
.shstrtab,0,-1
.debug_line,0,-16
[Unmapped],0,-16
.strtab,0,-87
.debug_str,0,-293

[github-actions]

Size increase report for "nrfconnect-example-build" from a62ee31

File	Section	File	VM
chip-shell.elf	text	72	72
chip-shell.elf	rodata	56	56
chip-shell.elf	device_handles	-8	-8
chip-lock.elf	text	72	72
chip-lock.elf	rodata	56	56
chip-lock.elf	device_handles	-8	-8

Full report output

BLOAT REPORT

Files found only in the build output:
    report.csv

Comparing ./master_artifact/chip-shell.elf and ./pull_artifact/chip-shell.elf:

sections,vmsize,filesize
.debug_info,0,992
.debug_loc,0,712
.debug_line,0,361
.debug_ranges,0,96
text,72,72
rodata,56,56
.debug_abbrev,0,55
.symtab,0,32
.debug_frame,0,16
.debug_aranges,0,8
.shstrtab,0,1
device_handles,-8,-8
.strtab,0,-77
.debug_str,0,-248

Comparing ./master_artifact/chip-lock.elf and ./pull_artifact/chip-lock.elf:

sections,vmsize,filesize
.debug_info,0,2058
.debug_loc,0,750
.debug_line,0,123
.debug_ranges,0,96
text,72,72
rodata,56,56
.debug_abbrev,0,55
.symtab,0,32
.debug_frame,0,16
.debug_aranges,0,8
.shstrtab,0,-1
device_handles,-8,-8
.strtab,0,-87
.debug_str,0,-258

[github-actions]

Size increase report for "esp32-example-build" from a62ee31

File	Section	File	VM
chip-all-clusters-app.elf	.flash.text	84	84
chip-all-clusters-app.elf	.flash.rodata	56	56

Full report output

BLOAT REPORT

Files found only in the build output:
    report.csv

Comparing ./master_artifact/chip-all-clusters-app.elf and ./pull_artifact/chip-all-clusters-app.elf:

sections,vmsize,filesize
.debug_info,0,2691
.debug_loc,0,121
.flash.text,84,84
.flash.rodata,56,56
.symtab,0,32
.debug_frame,0,24
.debug_aranges,0,8
.debug_abbrev,0,4
.shstrtab,0,1
.riscv.attributes,0,-1
.strtab,0,-9
.debug_line,0,-107
[Unmapped],0,-140
.debug_ranges,0,-160
.debug_str,0,-264

[kghost]

I feel that IsAckPending is indicating the active state of an exchange. It means that the exchange is busy generating an response, and it is unable to process incoming requests now.

HasPiggybackAckPending means that a ack is set, it actually means that the exchange have received a packet from peer. After this PR, all packets except the very beginning one will contains a valid ack.

So I just feel that these function names are a little bit confused.

[bzbarsky-apple]

@kghost Please suggest what you'd prefer as names?