Fix a panic when attempting to log when certificate should not be renewed

[antoniomika]

In v0.17.1, we're seeing the following panic occur (and subsequently, no recovery) over at sish.

This occurs because we don't use zap for logging, which leads to loggerNamed returning nil. I went through and checked all other logger references and made sure they are guarded by the same issue by checking that log is not nil.

panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x0 pc=0xa1f3ca]

goroutine 537 [running]:
go.uber.org/zap.(*Logger).check(0x0, 0x2, {0xc8e025, 0x22})
	/go/pkg/mod/go.uber.org/zap@v1.23.0/logger.go:291 +0x6a
go.uber.org/zap.(*Logger).Error(0xc0000ee038?, {0xc8e025?, 0x4?}, {0xc00071d300, 0x1, 0x1})
	/go/pkg/mod/go.uber.org/zap@v1.23.0/logger.go:228 +0x3e
github.com/caddyserver/certmagic.(*Config).renewDynamicCertificate.func3({0xdbb248, 0xc001529da0}, 0xc00011acc0?)
	/go/pkg/mod/github.com/caddyserver/certmagic@v0.17.1/handshake.go:655 +0xbb7
created by github.com/caddyserver/certmagic.(*Config).renewDynamicCertificate
	/go/pkg/mod/github.com/caddyserver/certmagic@v0.17.1/handshake.go:699 +0x1293

[mholt]

Aha, thank you.

What I might actually do soon is ensure a default log is always set... and then remove all the nil checks.