----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: http://reviews.gem5.org/r/2743/#review6052 -----------------------------------------------------------
I like the restructuring... I agree the micro-op loop is probably not realistic. Is there a reason to code a loop in C though, as opposed to just using '/' and '%'? - Steve Reinhardt On April 17, 2015, 1:32 p.m., Nilay Vaish wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > http://reviews.gem5.org/r/2743/ > ----------------------------------------------------------- > > (Updated April 17, 2015, 1:32 p.m.) > > > Review request for Default. > > > Repository: gem5 > > > Description > ------- > > Changeset 10790:25e772c6cb98 > --------------------------- > x86: reimplement integer division > > This patch reimplements integer division for x86 architecture. Earlier the > division algorithm was implemented using microops. I don't think that's what > happens in actual implementations. As I have read, there is separate division > unit that would carry out the whole operation. This means we only need to > issue a small number of microops, not a microop for each bit of the dividend, > as is happening currently. This patch implements the division algorithm in a > using four microops. The first microop reads the higher part of the dividend > and the divisor and records them in local registers. The second microop reads > the lower part of the dividend and carries out the division algorithm. The > next two microops populate the remainder and quotient registers (unchanged > from > before). > > > Diffs > ----- > > src/arch/x86/isa/insts/general_purpose/arithmetic/multiply_and_divide.py > d1df075f3b71 > src/arch/x86/isa/microops/regop.isa d1df075f3b71 > > Diff: http://reviews.gem5.org/r/2743/diff/ > > > Testing > ------- > > > Thanks, > > Nilay Vaish > > _______________________________________________ gem5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/gem5-dev
