You need to find concurrent conflict serializable schedule which means order of all conflicts(ww/wr/rw) between t1 and t2 must be in order t1->t2
For eg : i may wish to executed r2(b) just after r1(a) in order to achieve concurrency instead of waiting for t1 to finish
but you have to take care of given condition as i mentioned in first line
Now if u put r2(b) anywhere before w1(b) you will see there will be a conflict from T2 to T1 which is not allowed
so only concurrent schedule possible is
X:r1(A)w1(A)r1(B)w1(B)r2(B)w2(B)r2(A)w2(A)
which is a serial schedule :D
If u are still confused lemme know :)
