JetCalibTools: fix overflow/underflow in JMSCorrection

JetCalibTools: fix overflow/underflow in JMSCorrection

Hello,

This MR is fixing the treatment of underflows and overflows in JMSCorrection. The problem was reported in ATEAM-955.
Somehow in the trigger, it happens that we pass `pT_uncorr=3000` to `getMassCorr`, which corresponds precisely to `pTMax`.
In that case, ` if ( pT_uncorr > pTMax ) pT_uncorr = pTMax-1e-6 ;` does nothing, and `m_respFactorsMass[etabin]->Interpolate( pT_uncorr, mass_uncorr );` complains that we're out of the validity range.
So I changed `>` to `>=` and `<` to `<=`.

Cheers,
Bertrand