fix: divide by zero fatal crash for reprovide rate check

[dbeal-eth]

While I was running kubo on my instance, I was encountering somewhat frequent crashes of the process with the following stack trace:

2024-04-25 20:41:17.361	
2024-04-25T11:41:17.331332+00:00 cannon-repo-snap-154780564-4013d1 systemd[1]: ipfs.service: Consumed 5min 27.300s CPU time.
2024-04-25 20:41:17.361	
2024-04-25T11:41:17.331245+00:00 cannon-repo-snap-154780564-4013d1 systemd[1]: ipfs.service: Failed with result 'exit-code'.
2024-04-25 20:41:17.361	
2024-04-25T11:41:17.330698+00:00 cannon-repo-snap-154780564-4013d1 systemd[1]: ipfs.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
2024-04-25 20:41:17.361	
2024-04-25T11:41:17.296960+00:00 cannon-repo-snap-154780564-4013d1 ipfs[23032]: #011github.com/ipfs/boxo@v0.13.1/provider/reprovider.go:227 +0x88
2024-04-25 20:41:17.361	
2024-04-25T11:41:17.296912+00:00 cannon-repo-snap-154780564-4013d1 ipfs[23032]: created by github.com/ipfs/boxo/provider.(*reprovider).run in goroutine 1
2024-04-25 20:41:17.361	
repo-sgp1                |                         | 2024-04-25T11:41:17.296775+00:00 cannon-repo-snap-154780564-4013d1 ipfs[23032]: #011github.com/ipfs/boxo@v0.13.1/provider/reprovider.go:373 +0x9f3

2024-04-25 20:41:17.361	
2024-04-25T11:41:17.296775+00:00 cannon-repo-snap-154780564-4013d1 ipfs[23032]: #011github.com/ipfs/boxo@v0.13.1/provider/reprovider.go:373 +0x9f3

2024-04-25 20:41:17.361	
2024-04-25T11:41:17.296741+00:00 cannon-repo-snap-154780564-4013d1 ipfs[23032]: github.com/ipfs/boxo/provider.(*reprovider).run.func1()
2024-04-25 20:41:17.361	
2024-04-25T11:41:17.296704+00:00 cannon-repo-snap-154780564-4013d1 ipfs[23032]: #011github.com/ipfs/kubo@v0.23.0/core/node/provider.go:86 +0x535
2024-04-25 20:41:17.361	
2024-04-25T11:41:17.296656+00:00 cannon-repo-snap-154780564-4013d1 ipfs[23032]: github.com/ipfs/kubo/core/node.ProviderSys.func1.1(0x0?, 0x0?, 0x80, 0x0?)
2024-04-25 20:41:17.361	
2024-04-25T11:41:17.296583+00:00 cannon-repo-snap-154780564-4013d1 ipfs[23032]: goroutine 677 [running]:
2024-04-25 20:41:17.361	
2024-04-25T11:41:17.295863+00:00 cannon-repo-snap-154780564-4013d1 ipfs[23032]: panic: runtime error: integer divide by zero

I followed the stack trace to find the area where divide by 0 occurs. The solution seemed simple enough, so I went ahead and opened this PR with a simple change. I'm guessing that having a 0 count for reprovide is pretty bad, but will deal with that separately.

BTW I am not a go dev. My apologies if my solution seems too simplistic or a better way exists.

BTW it wasnt clear to me how/where I should update the changelog for this minor change. Please lmk. Thanks!

[dbeal-eth]

@hacdias would be super happy if you or someone on your team could take a look at this soon, thanks!

[hacdias]

I think this is a reasonable change. I see why count could be 0, so it makes sense to me. I don't think a changelog entry is necessary for such fix.