/* * arch/alpha/lib/stxncpy.S * Contributed by Richard Henderson (rth@tamu.edu) * * Copy no more than COUNT bytes of the null-terminated string from * SRC to DST. * * This is an internal routine used by strncpy, stpncpy, and strncat. * As such, it uses special linkage conventions to make implementation * of these public functions more efficient. * * On input: * t9 = return address * a0 = DST * a1 = SRC * a2 = COUNT * * Furthermore, COUNT may not be zero. * * On output: * t0 = last word written * t10 = bitmask (with one bit set) indicating the byte position of * the end of the range specified by COUNT * t12 = bitmask (with one bit set) indicating the last byte written * a0 = unaligned address of the last *word* written * a2 = the number of full words left in COUNT * * Furthermore, v0, a3-a5, t11, t12, and $at are untouched. */ #include .set noat .set noreorder .text /* There is a problem with either gdb (as of 4.16) or gas (as of 2.7) that doesn't like putting the entry point for a procedure somewhere in the middle of the procedure descriptor. Work around this by putting the aligned copy in its own procedure descriptor */ .ent stxncpy_aligned .align 3 stxncpy_aligned: .frame sp, 0, t9, 0 .prologue 0 /* On entry to this basic block: t0 == the first destination word for masking back in t1 == the first source word. */ /* Create the 1st output word and detect 0's in the 1st input word. */ lda t2, -1 # e1 : build a mask against false zero mskqh t2, a1, t2 # e0 : detection in the src word mskqh t1, a1, t3 # e0 : ornot t1, t2, t2 # .. e1 : mskql t0, a1, t0 # e0 : assemble the first output word cmpbge zero, t2, t8 # .. e1 : bits set iff null found or t0, t3, t0 # e0 : beq a2, $a_eoc # .. e1 : bne t8, $a_eos # .. e1 : /* On entry to this basic block: t0 == a source word not containing a null. */ $a_loop: stq_u t0, 0(a0) # e0 : addq a0, 8, a0 # .. e1 : ldq_u t0, 0(a1) # e0 : addq a1, 8, a1 # .. e1 : subq a2, 1, a2 # e0 : cmpbge zero, t0, t8 # .. e1 (stall) beq a2, $a_eoc # e1 : beq t8, $a_loop # e1 : /* Take care of the final (partial) word store. At this point the end-of-count bit is set in t8 iff it applies. On entry to this basic block we have: t0 == the source word containing the null t8 == the cmpbge mask that found it. */ $a_eos: negq t8, t12 # e0 : find low bit set and t8, t12, t12 # e1 (stall) /* For the sake of the cache, don't read a destination word if we're not going to need it. */ and t12, 0x80, t6 # e0 : bne t6, 1f # .. e1 (zdb) /* We're doing a partial word store and so need to combine our source and original destination words. */ ldq_u t1, 0(a0) # e0 : subq t12, 1, t6 # .. e1 : or t12, t6, t8 # e0 : unop # zapnot t0, t8, t0 # e0 : clear src bytes > null zap t1, t8, t1 # .. e1 : clear dst bytes <= null or t0, t1, t0 # e1 : 1: stq_u t0, 0(a0) # e0 : ret (t9) # e1 : /* Add the end-of-count bit to the eos detection bitmask. */ $a_eoc: or t10, t8, t8 br $a_eos .end stxncpy_aligned .align 3 .ent __stxncpy .globl __stxncpy __stxncpy: .frame sp, 0, t9, 0 .prologue 0 /* Are source and destination co-aligned? */ xor a0, a1, t1 # e0 : and a0, 7, t0 # .. e1 : find dest misalignment and t1, 7, t1 # e0 : addq a2, t0, a2 # .. e1 : bias count by dest misalignment subq a2, 1, a2 # e0 : and a2, 7, t2 # e1 : srl a2, 3, a2 # e0 : a2 = loop counter = (count - 1)/8 addq zero, 1, t10 # .. e1 : sll t10, t2, t10 # e0 : t10 = bitmask of last count byte bne t1, $unaligned # .. e1 : /* We are co-aligned; take care of a partial first word. */ ldq_u t1, 0(a1) # e0 : load first src word addq a1, 8, a1 # .. e1 : beq t0, stxncpy_aligned # avoid loading dest word if not needed ldq_u t0, 0(a0) # e0 : br stxncpy_aligned # .. e1 : /* The source and destination are not co-aligned. Align the destination and cope. We have to be very careful about not reading too much and causing a SEGV. */ .align 3 $u_head: /* We know just enough now to be able to assemble the first full source word. We can still find a zero at the end of it that prevents us from outputting the whole thing. On entry to this basic block: t0 == the first dest word, unmasked t1 == the shifted low bits of the first source word t6 == bytemask that is -1 in dest word bytes */ ldq_u t2, 8(a1) # e0 : load second src word addq a1, 8, a1 # .. e1 : mskql t0, a0, t0 # e0 : mask trailing garbage in dst extqh t2, a1, t4 # e0 : or t1, t4, t1 # e1 : first aligned src word complete mskqh t1, a0, t1 # e0 : mask leading garbage in src or t0, t1, t0 # e0 : first output word complete or t0, t6, t6 # e1 : mask original data for zero test cmpbge zero, t6, t8 # e0 : beq a2, $u_eocfin # .. e1 : bne t8, $u_final # e1 : lda t6, -1 # e1 : mask out the bits we have mskql t6, a1, t6 # e0 : already seen stq_u t0, 0(a0) # e0 : store first output word or t6, t2, t2 # .. e1 : cmpbge zero, t2, t8 # e0 : find nulls in second partial addq a0, 8, a0 # .. e1 : subq a2, 1, a2 # e0 : bne t8, $u_late_head_exit # .. e1 : /* Finally, we've got all the stupid leading edge cases taken care of and we can set up to enter the main loop. */ extql t2, a1, t1 # e0 : position hi-bits of lo word ldq_u t2, 8(a1) # .. e1 : read next high-order source word addq a1, 8, a1 # e0 : cmpbge zero, t2, t8 # e1 (stall) beq a2, $u_eoc # e1 : bne t8, $u_eos # e1 : /* Unaligned copy main loop. In order to avoid reading too much, the loop is structured to detect zeros in aligned source words. This has, unfortunately, effectively pulled half of a loop iteration out into the head and half into the tail, but it does prevent nastiness from accumulating in the very thing we want to run as fast as possible. On entry to this basic block: t1 == the shifted high-order bits from the previous source word t2 == the unshifted current source word We further know that t2 does not contain a null terminator. */ .align 3 $u_loop: extqh t2, a1, t0 # e0 : extract high bits for current word addq a1, 8, a1 # .. e1 : extql t2, a1, t3 # e0 : extract low bits for next time addq a0, 8, a0 # .. e1 : or t0, t1, t0 # e0 : current dst word now complete ldq_u t2, 0(a1) # .. e1 : load high word for next time stq_u t0, -8(a0) # e0 : save the current word mov t3, t1 # .. e1 : subq a2, 1, a2 # e0 : cmpbge zero, t2, t8 # .. e1 : test new word for eos beq a2, $u_eoc # e1 : beq t8, $u_loop # e1 : /* We've found a zero somewhere in the source word we just read. If it resides in the lower half, we have one (probably partial) word to write out, and if it resides in the upper half, we have one full and one partial word left to write out. On entry to this basic block: t1 == the shifted high-order bits from the previous source word t2 == the unshifted current source word. */ $u_eos: extqh t2, a1, t0 # e0 : or t0, t1, t0 # e1 : first (partial) source word complete cmpbge zero, t0, t8 # e0 : is the null in this first bit? bne t8, $u_final # .. e1 (zdb) stq_u t0, 0(a0) # e0 : the null was in the high-order bits addq a0, 8, a0 # .. e1 : subq a2, 1, a2 # e1 : $u_late_head_exit: extql t2, a1, t0 # .. e0 : cmpbge zero, t0, t8 # e0 : or t8, t10, t6 # e1 : cmoveq a2, t6, t8 # e0 : nop # .. e1 : /* Take care of a final (probably partial) result word. On entry to this basic block: t0 == assembled source word t8 == cmpbge mask that found the null. */ $u_final: negq t8, t6 # e0 : isolate low bit set and t6, t8, t12 # e1 : and t12, 0x80, t6 # e0 : avoid dest word load if we can bne t6, 1f # .. e1 (zdb) ldq_u t1, 0(a0) # e0 : subq t12, 1, t6 # .. e1 : or t6, t12, t8 # e0 : zapnot t0, t8, t0 # .. e1 : kill source bytes > null zap t1, t8, t1 # e0 : kill dest bytes <= null or t0, t1, t0 # e1 : 1: stq_u t0, 0(a0) # e0 : ret (t9) # .. e1 : $u_eoc: # end-of-count extqh t2, a1, t0 or t0, t1, t0 cmpbge zero, t0, t8 $u_eocfin: # end-of-count, final word or t10, t8, t8 br $u_final /* Unaligned copy entry point. */ .align 3 $unaligned: ldq_u t1, 0(a1) # e0 : load first source word and a0, 7, t4 # .. e1 : find dest misalignment and a1, 7, t5 # e0 : find src misalignment /* Conditionally load the first destination word and a bytemask with 0xff indicating that the destination byte is sacrosanct. */ mov zero, t0 # .. e1 : mov zero, t6 # e0 : beq t4, 1f # .. e1 : ldq_u t0, 0(a0) # e0 : lda t6, -1 # .. e1 : mskql t6, a0, t6 # e0 : 1: subq a1, t4, a1 # .. e1 : sub dest misalignment from src addr /* If source misalignment is larger than dest misalignment, we need extra startup checks to avoid SEGV. */ cmplt t4, t5, t12 # e1 : extql t1, a1, t1 # .. e0 : shift src into place lda t2, -1 # e0 : for creating masks later beq t12, $u_head # e1 : mskqh t2, t5, t2 # e0 : begin src byte validity mask cmpbge zero, t1, t8 # .. e1 : is there a zero? extql t2, a1, t2 # e0 : or t8, t10, t5 # .. e1 : test for end-of-count too cmpbge zero, t2, t3 # e0 : cmoveq a2, t5, t8 # .. e1 : andnot t8, t3, t8 # e0 : beq t8, $u_head # .. e1 (zdb) /* At this point we've found a zero in the first partial word of the source. We need to isolate the valid source data and mask it into the original destination data. (Incidentally, we know that we'll need at least one byte of that original dest word.) */ ldq_u t0, 0(a0) # e0 : negq t8, t6 # .. e1 : build bitmask of bytes <= zero mskqh t1, t4, t1 # e0 : and t6, t8, t12 # .. e1 : subq t12, 1, t6 # e0 : or t6, t12, t8 # e1 : zapnot t2, t8, t2 # e0 : prepare source word; mirror changes zapnot t1, t8, t1 # .. e1 : to source validity mask andnot t0, t2, t0 # e0 : zero place for source to reside or t0, t1, t0 # e1 : and put it there stq_u t0, 0(a0) # e0 : ret (t9) # .. e1 : .end __stxncpy