[dns-sd] fix SRP service name length

[wgtdkp]

Problem

The SRP service name buffers are too small.

Summary of Changes

Fix SRP service & host name buffer issue.

[andy31415]

On this PR I see the 'run unit tests' stuck in the run state (without output unfortunately) for 50 minutes. Unsure if PR-specific or a flake. Given that every one of the runners does this, it may be something in the change itself.

[andy31415]

I cancelled the workflows with the intent of restarting them, however I do not seem to get a 'restart all' button yet.
If I am unable to do so, @wgtdkp could you push a minor update to kick the presubmits again?

[github-actions]

Size increase report for "esp32-example-build" from 8185019

File	Section	File	VM
chip-all-clusters-app.elf	.flash.rodata	-8	-8

Full report output

BLOAT REPORT

Files found only in the build output:
    report.csv

Comparing ./master_artifact/chip-pigweed-app.elf and ./pull_artifact/chip-pigweed-app.elf:

sections,vmsize,filesize

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

sections,vmsize,filesize
[Unmapped],0,8
.flash.rodata,-8,-8

[github-actions]

Size increase report for "nrfconnect-example-build" from 8185019

File	Section	File	VM
chip-lighting.elf	rodata	56	52
chip-lighting.elf	text	40	40
chip-lighting.elf	device_handles	-8	-8
chip-lighting.elf	bss	0	-13
chip-lighting.elf	[LOAD #3 [RW]]	0	-19
chip-shell.elf	[LOAD #3 [RW]]	0	31
chip-shell.elf	bss	0	-31
chip-lock.elf	rodata	56	52
chip-lock.elf	text	36	36
chip-lock.elf	device_handles	-4	-4
chip-lock.elf	bss	0	-9
chip-lock.elf	[LOAD #3 [RW]]	0	-23

Full report output

BLOAT REPORT

Files found only in the build output:
    report.csv

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

sections,vmsize,filesize
.debug_info,0,1497
rodata,52,56
text,40,40
.symtab,0,32
.debug_str,0,28
.debug_frame,0,12
.debug_line,0,3
.debug_loc,0,-4
device_handles,-8,-8
bss,-13,0
[LOAD #3 [RW]],-19,0

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

sections,vmsize,filesize
.debug_info,0,481
[LOAD #3 [RW]],31,0
.debug_str,0,28
.debug_frame,0,12
.debug_loc,0,4
.debug_line,0,3
bss,-31,0

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

sections,vmsize,filesize
.debug_info,0,777
rodata,52,56
text,36,36
.symtab,0,32
.debug_str,0,28
.debug_frame,0,12
.debug_line,0,3
.debug_loc,0,-4
device_handles,-4,-4
bss,-9,0
[LOAD #3 [RW]],-23,0

[woody-apple]

I kicked the Tests again, for some reason they're getting stuck...

[andy31415]

I kicked the Tests again, for some reason they're getting stuck...

I am totally puzzled here - I have never seen these tests hang on any other PR, however I am unclear on why this PR would cause system tests to misbehave either (and across all compilers to boot). I did copy the PR and tried locally, it passes (except I found a Mdns hang which I fixed in #6721). Kangping reported his workstation hangs on TestSystemObject but I could not see that.

[wgtdkp]

@woody-apple @andy31415 Thanks for helping resolve the test hang issue. I am sorry that it is too later for me to fix this issue last night, but it seems the hang issue is caused by the failure of the TestMdns test case and it is my change in this PR that is failing the TestMdns case :(

I submitted #6771 to fix the issue.