Hallo Nagarajan,

topology of port drivers is a little bit complicated. At some port pins there are up to three active pull-ups. A strong, a weak and a very weak. This, in order to achieve input and output capability.
Strong pull-up is activated when there is need for a toggling from low to high. For only two oscillator periods this strong pull-up is activated, for being able to charge any capacitive load. Afterwards, strong pull-up is switched-off and only weak pull-up is keeping the output in high state.
This topology is needed for having both, fast edges and the ability to tolerate 'short circuits' at port pin, when it is used as input. Would port pin be driven permanently with strong pull-up there would be a damage.

Of course, activation of strong pull-up will increase supply current, but only when there is a low to high change and when some capacitive load has to be charged. When there is no change necessary increase of supply current will be neglectable.

Port3 pins sometimes suffer from removing strong pull-up after driving edge. RD# and WR# for instance. After strong pull-up is switched-off, only weak pull-up is driving line and impedance of driver output increases to some kOhms. When there is any noise capacitively coupled on this line, increased driver impedance can be to high too suppress it enough, like a classical HCMOS, e.g. Experience showed, that adding an additional resistor pull-up sometimes can help to assure signal integritiy.

Keep in mind, that Port0 outputs are quite different. Mostly there are open drain outpus which can float to any potential, when fets are switched off. Then, you must have some pull-up.

Bye,
Kai