tail: Fix heap overflow in -F case.

[des]

Summary

The number of events we track can vary over time, but we only allocated
enough space for the exact number of events we are tracking when we
first begin, resulting in a trivially reproducable heap overflow. Fix
this by allocating enough space for the greatest possible number of
events (two per file) and clean up the code a bit.

Also add a test case which triggers the aforementioned heap overflow,
although we don't currently have a way to detect it.

MFC after: 1 week
Sponsored by: Klara, Inc.

[allanjude]

reviewed-by: allanjude

[markj]

It is silly that kq and ev are global variables instead of being passed as parameters to set_events() (and action should be a return value of set_events()), but fixing that is out of the scope of this patch.

Also add a test case which triggers the aforementioned heap overflow, although we don't currently have a way to detect it.

Does ASAN detect it?

usr.bin/tail/forward.c

338

since you're changing this line anyway.

[des]

In D42839#977069, @markj wrote:

Also add a test case which triggers the aforementioned heap overflow, although we don't currently have a way to detect it.

Does ASAN detect it?

I haven't tried, but I would be surprised if it didn't. Valgrind loses its mind, just try valgrind tail -F /tmp/nonexistent in one terminal and then touch /tmp/nonexistent in another.

[des]

usr.bin/tail/forward.c

338

There's at least one other unrelated case in another file so I'll do them separately.