src,dns: reserve vector capacity to avoid unnecessary reallocations

[agape1225]

Message::Serialize() and ParseTxtReply() both grow a LocalVector
via repeated push_back() without reserving capacity first, even
though the upper bound on the number of elements is known before the
loop starts. This adds reserve() in both places, following the
usual idiom of reserving upfront when the size (or a bound on it) is
already known, to avoid the repeated allocate/copy/free cycle that
happens each time capacity is exceeded.

Note: in Message::Serialize, transfer_list_v.length() is an upper
bound rather than an exact count (only ArrayBuffer entries get
pushed into array_buffers), so a transfer list with few or no
ArrayBuffers reserves capacity that ends up unused. No functional
change either way.

[nodejs-github-bot]

Review requested:

• @nodejs/net

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 90.13%. Comparing base (4a5eb1c) to head (403b55e).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #64763      +/-   ##
==========================================
- Coverage   90.15%   90.13%   -0.02%     
==========================================
  Files         744      744              
  Lines      242517   242519       +2     
  Branches    45688    45695       +7     
==========================================
- Hits       218642   218600      -42     
- Misses      15358    15410      +52     
+ Partials     8517     8509       -8     

Files with missing lines	Coverage Δ
src/cares_wrap.cc	62.93% <100.00%> (+0.02%)	⬆️
src/node_messaging.cc	82.88% <100.00%> (+0.01%)	⬆️

... and 26 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.