Re: [PATCH 1/2] telemetry: use malloc instead of variable length array

[Bruce Richardson <bruce.richardson@intel.com>]

On Tue, Apr 04, 2023 at 06:20:04PM -0700, Stephen Hemminger wrote:
> On Tue, 4 Apr 2023 10:34:01 -0700
> Tyler Retzlaff <roretzla@linux.microsoft.com> wrote:
> 
> > > > > I think suggestion #2 above should cover most cases, in which case using
> > > > > your original suggestion of malloc would be ok too for the rare case (if
> > > > > ever) where we don't just have one terminator on the end.  
> > > > 
> > > > maybe a dumb'd down compromise is to have a fixed stack limit and then
> > > > if it is exceeded always just go to malloc/free?
> > > >   
> > > Perhaps. If you like, I have have a try at implementing my own suggestions
> > > above tomorrow. I'd like if we can get the "single-character-saving" option
> > > working, because that would be the most efficient method of all.  
> > 
> > that would be great, i'll take the help i can get.
> 
> 
> For the json print library in iproute2, it streams output using
> stdio, which avoids lots of string processing issues. Well, it moves them
> to be standard library.

Yes, but it does have a number of downsides:

* streaming means that we have to worry about message boundaries when
  transferring across sockets like in the telemetry case. With telemetry we
  have a packet-based (message based) interface, so we need to buffer all
  output to a string anyway. If we ever add a streaming (socket or fifo)
  interface, then the iproute2 implementation may be a better fit.

* Last I looked at the iproute2 json implementation, it requires the user to
  track how many outstanding closing brackets of what type are needed,
  meaning that you have invalid json at intermediate stages, and makes it
  awkward if you run out of space in a buffer, i.e. no room for terminators.
  The implementation here, on the other hand, ensures that the output is
  always valid json, so for example, when adding a second string to an
  array of strings we go from ["str1"] to ["str1","str2"] directly - or in
  the case of insufficient space we leave the original version unmodified.
  For working with the case where we have a fixed buffer, this is a massive
  benefit.

/Bruce