misleading documentation about ValidatorSigningInfo.StartHeight

[0Tech]

A comment in the proto file says start_height would be set to the height at which the validator was unjailed:

cosmos-sdk/proto/cosmos/slashing/v1beta1/slashing.proto

Lines 19 to 20 in 1c468de

	// Height at which validator was first a candidate OR was unjailed
	int64 start_height = 2;

But this is not true. You can see the counter example in the test code:

cosmos-sdk/x/slashing/keeper/keeper_test.go

Lines 263 to 281 in 1c468de

	// validator rejoins and starts signing again
	app.StakingKeeper.Unjail(ctx, consAddr)
	app.SlashingKeeper.HandleValidatorSignature(ctx, val.Address(), newPower, true)
	height++
	// validator should not be kicked since we reset counter/array when it was jailed
	staking.EndBlocker(ctx, app.StakingKeeper)
	tstaking.CheckValidator(valAddr, stakingtypes.Bonded, false)
	// validator misses 501 blocks
	latest = height
	for ; height < latest+501; height++ {
	ctx = ctx.WithBlockHeight(height)
	app.SlashingKeeper.HandleValidatorSignature(ctx, val.Address(), newPower, false)
	}
	// validator should now be jailed & kicked
	staking.EndBlocker(ctx, app.StakingKeeper)
	tstaking.CheckValidator(valAddr, stakingtypes.Unbonding, true)

Which means the test expects that the validator would be punished even if another 1000 blocks had not passed since the unjail.

I think at least one of the three parts should be fixed - the comment, the test code or the behavior.

Please correct me if there is anything I have missed.