| Message ID | patch-18113-tamar@arm.com |
|---|---|
| State | New |
| Headers | show |
| Series | middle-end: Fix dominators updates when peeling with multiple exits [PR113144] | expand |
On Fri, 29 Dec 2023, Tamar Christina wrote: > Hi All, > > Only trying to update certain dominators doesn't seem to work very well > because as the loop gets versioned, peeled, or skip_vector then we end up with > very complicated control flow. This means that the final merge blocks for the > loop exit are not easy to find or update. > > Instead of trying to pick which exits to update, this changes it to update all > the blocks reachable by the new exits. This is because they'll contain common > blocks with e.g. the versioned loop. It's these blocks that need an update > most of the time. > > Bootstrapped Regtested on aarch64-none-linux-gnu and no issues. > > Ok for master? This makes it quadratic in the number of vectorized early exit loops in a function. The vectorizer CFG manipulation operates in a local enough bubble that programmatic updating of dominators should be possible (after all we manage to produce correct SSA form!), the proposed change gets us too far off to a point where re-computating dominance info is likely cheaper (but no, we shouldn't do this either). Can you instead give manual updating a try again? I think versioning should produce up-to-date dominator info, it's only when you redirect branches during peeling that you'd need adjustments - but IIRC we're never introducing new merges? IIRC we can't wipe dominators during transform since we query them during code generation. We possibly could code generate all CFG manipulations of all vectorized loops, recompute all dominators and then do code generation of all vectorized loops. But then we're doing a loop transform and the exits will ultimatively end up in the same place, so the CFG and dominator update is bound to where the original exits went to. Richard > Thanks, > Tamar > > gcc/ChangeLog: > > PR middle-end/113144 > * tree-vect-loop-manip.cc (slpeel_tree_duplicate_loop_to_edge_cfg): > Update all dominators reachable from exit. > > gcc/testsuite/ChangeLog: > > PR middle-end/113144 > * gcc.dg/vect/vect-early-break_94-pr113144.c: New test. > > --- inline copy of patch -- > diff --git a/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > new file mode 100644 > index 0000000000000000000000000000000000000000..903fe7be6621e81db6f29441e4309fa213d027c5 > --- /dev/null > +++ b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > @@ -0,0 +1,41 @@ > +/* { dg-do compile } */ > +/* { dg-add-options vect_early_break } */ > +/* { dg-require-effective-target vect_early_break } */ > +/* { dg-require-effective-target vect_int } */ > + > +/* { dg-final { scan-tree-dump "LOOP VECTORIZED" "vect" } } */ > + > +long tar_atol256_max, tar_atol256_size, tar_atosl_min; > +char tar_atol256_s; > +void __errno_location(); > + > + > +inline static long tar_atol256(long min) { > + char c; > + int sign; > + c = tar_atol256_s; > + sign = c; > + while (tar_atol256_size) { > + if (c != sign) > + return sign ? min : tar_atol256_max; > + c = tar_atol256_size--; > + } > + if ((c & 128) != (sign & 128)) > + return sign ? min : tar_atol256_max; > + return 0; > +} > + > +inline static long tar_atol(long min) { > + return tar_atol256(min); > +} > + > +long tar_atosl() { > + long n = tar_atol(-1); > + if (tar_atosl_min) { > + __errno_location(); > + return 0; > + } > + if (n > 0) > + return 0; > + return n; > +} > diff --git a/gcc/tree-vect-loop-manip.cc b/gcc/tree-vect-loop-manip.cc > index 1066ea17c5674e03412b3dcd8a62ddf4dd54cf31..3810983a80c8b989be9fd9a9993642069fd39b99 100644 > --- a/gcc/tree-vect-loop-manip.cc > +++ b/gcc/tree-vect-loop-manip.cc > @@ -1716,8 +1716,6 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class loop *loop, edge loop_exit, > /* Now link the alternative exits. */ > if (multiple_exits_p) > { > - set_immediate_dominator (CDI_DOMINATORS, new_preheader, > - main_loop_exit_block); > for (auto gsi_from = gsi_start_phis (loop->header), > gsi_to = gsi_start_phis (new_preheader); > !gsi_end_p (gsi_from) && !gsi_end_p (gsi_to); > @@ -1751,12 +1749,26 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class loop *loop, edge loop_exit, > > /* Finally after wiring the new epilogue we need to update its main exit > to the original function exit we recorded. Other exits are already > - correct. */ > + correct. Because of versioning, skip vectors and others we must update > + the dominators of every node reachable by the new exits. */ > if (multiple_exits_p) > { > update_loop = new_loop; > - for (edge e : get_loop_exit_edges (loop)) > - doms.safe_push (e->dest); > + hash_set <basic_block> visited; > + auto_vec <edge> workset; > + edge ev; > + edge_iterator ei; > + workset.safe_splice (get_loop_exit_edges (loop)); > + while (!workset.is_empty ()) > + { > + auto bb = workset.pop ()->dest; > + if (visited.add (bb)) > + continue; > + doms.safe_push (bb); > + FOR_EACH_EDGE (ev, ei, bb->succs) > + workset.safe_push (ev); > + } > + visited.empty (); > doms.safe_push (exit_dest); > > /* Likely a fall-through edge, so update if needed. */ > > > > >
> This makes it quadratic in the number of vectorized early exit loops > in a function. The vectorizer CFG manipulation operates in a local > enough bubble that programmatic updating of dominators should be > possible (after all we manage to produce correct SSA form!), the > proposed change gets us too far off to a point where re-computating > dominance info is likely cheaper (but no, we shouldn't do this either). > > Can you instead give manual updating a try again? I think > versioning should produce up-to-date dominator info, it's only > when you redirect branches during peeling that you'd need > adjustments - but IIRC we're never introducing new merges? > > IIRC we can't wipe dominators during transform since we query them > during code generation. We possibly could code generate all > CFG manipulations of all vectorized loops, recompute all dominators > and then do code generation of all vectorized loops. > > But then we're doing a loop transform and the exits will ultimatively > end up in the same place, so the CFG and dominator update is bound to > where the original exits went to. Yeah that's a fair point, the issue is specifically with at_exit. So how about: When we peel at_exit we are moving the new loop at the exit of the previous loop. This means that the blocks outside the loop dat the previous loop used to dominate are no longer being dominated by it. The new dominators however are hard to predict since if the loop has multiple exits and all the exits are an "early" one then we always execute the scalar loop. In this case the scalar loop can completely dominate the new loop. If we later have skip_vector then there's an additional skip edge added that might change the dominators. The previous patch would force an update of all blocks reachable from the new exits. This one updates *only* blocks that we know the scalar exits dominated. For the examples this reduces the blocks to update from 18 to 3. Bootstrapped Regtested on aarch64-none-linux-gnu, x86_64-pc-linux-gnu and no issues normally and with --enable-checking=release --enable-lto --with-build-config=bootstrap-O3 --enable-checking=yes,rtl,extra. Ok for master? Thanks, Tamar gcc/ChangeLog: PR tree-optimization/113144 PR tree-optimization/113145 * tree-vect-loop-manip.cc (slpeel_tree_duplicate_loop_to_edge_cfg): Update all BB that the original exits dominated. gcc/testsuite/ChangeLog: PR tree-optimization/113144 PR tree-optimization/113145 * gcc.dg/vect/vect-early-break_94-pr113144.c: New test. --- inline copy of patch --- diff --git a/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c new file mode 100644 index 0000000000000000000000000000000000000000..903fe7be6621e81db6f29441e4309fa213d027c5 --- /dev/null +++ b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c @@ -0,0 +1,41 @@ +/* { dg-do compile } */ +/* { dg-add-options vect_early_break } */ +/* { dg-require-effective-target vect_early_break } */ +/* { dg-require-effective-target vect_int } */ + +/* { dg-final { scan-tree-dump "LOOP VECTORIZED" "vect" } } */ + +long tar_atol256_max, tar_atol256_size, tar_atosl_min; +char tar_atol256_s; +void __errno_location(); + + +inline static long tar_atol256(long min) { + char c; + int sign; + c = tar_atol256_s; + sign = c; + while (tar_atol256_size) { + if (c != sign) + return sign ? min : tar_atol256_max; + c = tar_atol256_size--; + } + if ((c & 128) != (sign & 128)) + return sign ? min : tar_atol256_max; + return 0; +} + +inline static long tar_atol(long min) { + return tar_atol256(min); +} + +long tar_atosl() { + long n = tar_atol(-1); + if (tar_atosl_min) { + __errno_location(); + return 0; + } + if (n > 0) + return 0; + return n; +} diff --git a/gcc/tree-vect-loop-manip.cc b/gcc/tree-vect-loop-manip.cc index 76d4979c0b3b374dcaacf6825a95a8714114a63b..9bacaa182a3919cae1cb99dfc5ae4923e1f93376 100644 --- a/gcc/tree-vect-loop-manip.cc +++ b/gcc/tree-vect-loop-manip.cc @@ -1719,8 +1719,6 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class loop *loop, edge loop_exit, /* Now link the alternative exits. */ if (multiple_exits_p) { - set_immediate_dominator (CDI_DOMINATORS, new_preheader, - main_loop_exit_block); for (auto gsi_from = gsi_start_phis (loop->header), gsi_to = gsi_start_phis (new_preheader); !gsi_end_p (gsi_from) && !gsi_end_p (gsi_to); @@ -1776,7 +1774,14 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class loop *loop, edge loop_exit, { update_loop = new_loop; for (edge e : get_loop_exit_edges (loop)) - doms.safe_push (e->dest); + { + /* Basic blocks that the old loop dominated are now dominated by + the new loop and so we have to update those. */ + for (auto bb : get_all_dominated_blocks (CDI_DOMINATORS, e->src)) + if (!flow_bb_inside_loop_p (loop, bb)) + doms.safe_push (bb); + doms.safe_push (e->dest); + } doms.safe_push (exit_dest); /* Likely a fall-through edge, so update if needed. */
On Tue, 9 Jan 2024, Tamar Christina wrote: > > This makes it quadratic in the number of vectorized early exit loops > > in a function. The vectorizer CFG manipulation operates in a local > > enough bubble that programmatic updating of dominators should be > > possible (after all we manage to produce correct SSA form!), the > > proposed change gets us too far off to a point where re-computating > > dominance info is likely cheaper (but no, we shouldn't do this either). > > > > Can you instead give manual updating a try again? I think > > versioning should produce up-to-date dominator info, it's only > > when you redirect branches during peeling that you'd need > > adjustments - but IIRC we're never introducing new merges? > > > > IIRC we can't wipe dominators during transform since we query them > > during code generation. We possibly could code generate all > > CFG manipulations of all vectorized loops, recompute all dominators > > and then do code generation of all vectorized loops. > > > > But then we're doing a loop transform and the exits will ultimatively > > end up in the same place, so the CFG and dominator update is bound to > > where the original exits went to. > > Yeah that's a fair point, the issue is specifically with at_exit. So how about: > > When we peel at_exit we are moving the new loop at the exit of the previous > loop. This means that the blocks outside the loop dat the previous loop used to > dominate are no longer being dominated by it. Hmm, indeed. Note this does make the dominator update O(function-size) and when vectorizing multiple loops in a function this becomes quadratic. That's quite unfortunate so I wonder if we can delay the update to the parts we do not need up-to-date dominators during vectorization (of course it gets fragile with having only partly correct dominators). > The new dominators however are hard to predict since if the loop has multiple > exits and all the exits are an "early" one then we always execute the scalar > loop. In this case the scalar loop can completely dominate the new loop. > > If we later have skip_vector then there's an additional skip edge added that > might change the dominators. > > The previous patch would force an update of all blocks reachable from the new > exits. This one updates *only* blocks that we know the scalar exits dominated. > > For the examples this reduces the blocks to update from 18 to 3. > > Bootstrapped Regtested on aarch64-none-linux-gnu, x86_64-pc-linux-gnu > and no issues normally and with --enable-checking=release --enable-lto > --with-build-config=bootstrap-O3 --enable-checking=yes,rtl,extra. > > Ok for master? See below. > Thanks, > Tamar > > gcc/ChangeLog: > > PR tree-optimization/113144 > PR tree-optimization/113145 > * tree-vect-loop-manip.cc (slpeel_tree_duplicate_loop_to_edge_cfg): > Update all BB that the original exits dominated. > > gcc/testsuite/ChangeLog: > > PR tree-optimization/113144 > PR tree-optimization/113145 > * gcc.dg/vect/vect-early-break_94-pr113144.c: New test. > > --- inline copy of patch --- > > diff --git a/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > new file mode 100644 > index 0000000000000000000000000000000000000000..903fe7be6621e81db6f29441e4309fa213d027c5 > --- /dev/null > +++ b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > @@ -0,0 +1,41 @@ > +/* { dg-do compile } */ > +/* { dg-add-options vect_early_break } */ > +/* { dg-require-effective-target vect_early_break } */ > +/* { dg-require-effective-target vect_int } */ > + > +/* { dg-final { scan-tree-dump "LOOP VECTORIZED" "vect" } } */ > + > +long tar_atol256_max, tar_atol256_size, tar_atosl_min; > +char tar_atol256_s; > +void __errno_location(); > + > + > +inline static long tar_atol256(long min) { > + char c; > + int sign; > + c = tar_atol256_s; > + sign = c; > + while (tar_atol256_size) { > + if (c != sign) > + return sign ? min : tar_atol256_max; > + c = tar_atol256_size--; > + } > + if ((c & 128) != (sign & 128)) > + return sign ? min : tar_atol256_max; > + return 0; > +} > + > +inline static long tar_atol(long min) { > + return tar_atol256(min); > +} > + > +long tar_atosl() { > + long n = tar_atol(-1); > + if (tar_atosl_min) { > + __errno_location(); > + return 0; > + } > + if (n > 0) > + return 0; > + return n; > +} > diff --git a/gcc/tree-vect-loop-manip.cc b/gcc/tree-vect-loop-manip.cc > index 76d4979c0b3b374dcaacf6825a95a8714114a63b..9bacaa182a3919cae1cb99dfc5ae4923e1f93376 100644 > --- a/gcc/tree-vect-loop-manip.cc > +++ b/gcc/tree-vect-loop-manip.cc > @@ -1719,8 +1719,6 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class loop *loop, edge loop_exit, > /* Now link the alternative exits. */ > if (multiple_exits_p) > { > - set_immediate_dominator (CDI_DOMINATORS, new_preheader, > - main_loop_exit_block); > for (auto gsi_from = gsi_start_phis (loop->header), > gsi_to = gsi_start_phis (new_preheader); > !gsi_end_p (gsi_from) && !gsi_end_p (gsi_to); > @@ -1776,7 +1774,14 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class loop *loop, edge loop_exit, > { > update_loop = new_loop; > for (edge e : get_loop_exit_edges (loop)) > - doms.safe_push (e->dest); > + { > + /* Basic blocks that the old loop dominated are now dominated by > + the new loop and so we have to update those. */ > + for (auto bb : get_all_dominated_blocks (CDI_DOMINATORS, e->src)) > + if (!flow_bb_inside_loop_p (loop, bb)) > + doms.safe_push (bb); > + doms.safe_push (e->dest); > + } I think you'll get duplicate blocks that way. Maybe simplify this all by instead doing auto doms = get_all_dominated_blocks (CDI_DOMINATORS, loop->header); for (unsigned i = 0; i < doms.length (); ++i) if (flow_bb_inside_loop_p (loop, doms[i])) doms.unordered_remove (i); ? OK with that change, but really we should see to avoid this quadraticness :/ It's probably not too bad right now given we have quite some restrictions on vectorizing loops with multiple exits, but I suggest you try an artificial testcase with the "same" loop repeated N times to see whether dominance compute creeps up in the profile. Richard.
> -----Original Message----- > From: Richard Biener <rguenther@suse.de> > Sent: Tuesday, January 9, 2024 12:26 PM > To: Tamar Christina <Tamar.Christina@arm.com> > Cc: gcc-patches@gcc.gnu.org; nd <nd@arm.com>; jlaw@ventanamicro.com > Subject: RE: [PATCH]middle-end: Fix dominators updates when peeling with > multiple exits [PR113144] > > On Tue, 9 Jan 2024, Tamar Christina wrote: > > > > This makes it quadratic in the number of vectorized early exit loops > > > in a function. The vectorizer CFG manipulation operates in a local > > > enough bubble that programmatic updating of dominators should be > > > possible (after all we manage to produce correct SSA form!), the > > > proposed change gets us too far off to a point where re-computating > > > dominance info is likely cheaper (but no, we shouldn't do this either). > > > > > > Can you instead give manual updating a try again? I think > > > versioning should produce up-to-date dominator info, it's only > > > when you redirect branches during peeling that you'd need > > > adjustments - but IIRC we're never introducing new merges? > > > > > > IIRC we can't wipe dominators during transform since we query them > > > during code generation. We possibly could code generate all > > > CFG manipulations of all vectorized loops, recompute all dominators > > > and then do code generation of all vectorized loops. > > > > > > But then we're doing a loop transform and the exits will ultimatively > > > end up in the same place, so the CFG and dominator update is bound to > > > where the original exits went to. > > > > Yeah that's a fair point, the issue is specifically with at_exit. So how about: > > > > When we peel at_exit we are moving the new loop at the exit of the previous > > loop. This means that the blocks outside the loop dat the previous loop used to > > dominate are no longer being dominated by it. > > Hmm, indeed. Note this does make the dominator update O(function-size) > and when vectorizing multiple loops in a function this becomes > quadratic. That's quite unfortunate so I wonder if we can delay the > update to the parts we do not need up-to-date dominators during > vectorization (of course it gets fragile with having only partly > correct dominators). Fair, I created https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113290 and will tackle it when I add SLP support in GCC 15. I think the problem is, and the reason we do early dominator correction and validation is because the same function is used by loop distribution. But you're right that during vectorization we perform dominators update twice now. So Maybe we should have a parameter to indicate whether dominators should be updated? Thanks, Tamar > > > The new dominators however are hard to predict since if the loop has multiple > > exits and all the exits are an "early" one then we always execute the scalar > > loop. In this case the scalar loop can completely dominate the new loop. > > > > If we later have skip_vector then there's an additional skip edge added that > > might change the dominators. > > > > The previous patch would force an update of all blocks reachable from the new > > exits. This one updates *only* blocks that we know the scalar exits dominated. > > > > For the examples this reduces the blocks to update from 18 to 3. > > > > Bootstrapped Regtested on aarch64-none-linux-gnu, x86_64-pc-linux-gnu > > and no issues normally and with --enable-checking=release --enable-lto > > --with-build-config=bootstrap-O3 --enable-checking=yes,rtl,extra. > > > > Ok for master? > > See below. > > > Thanks, > > Tamar > > > > gcc/ChangeLog: > > > > PR tree-optimization/113144 > > PR tree-optimization/113145 > > * tree-vect-loop-manip.cc (slpeel_tree_duplicate_loop_to_edge_cfg): > > Update all BB that the original exits dominated. > > > > gcc/testsuite/ChangeLog: > > > > PR tree-optimization/113144 > > PR tree-optimization/113145 > > * gcc.dg/vect/vect-early-break_94-pr113144.c: New test. > > > > --- inline copy of patch --- > > > > diff --git a/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > > new file mode 100644 > > index > 0000000000000000000000000000000000000000..903fe7be6621e81db6f294 > 41e4309fa213d027c5 > > --- /dev/null > > +++ b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > > @@ -0,0 +1,41 @@ > > +/* { dg-do compile } */ > > +/* { dg-add-options vect_early_break } */ > > +/* { dg-require-effective-target vect_early_break } */ > > +/* { dg-require-effective-target vect_int } */ > > + > > +/* { dg-final { scan-tree-dump "LOOP VECTORIZED" "vect" } } */ > > + > > +long tar_atol256_max, tar_atol256_size, tar_atosl_min; > > +char tar_atol256_s; > > +void __errno_location(); > > + > > + > > +inline static long tar_atol256(long min) { > > + char c; > > + int sign; > > + c = tar_atol256_s; > > + sign = c; > > + while (tar_atol256_size) { > > + if (c != sign) > > + return sign ? min : tar_atol256_max; > > + c = tar_atol256_size--; > > + } > > + if ((c & 128) != (sign & 128)) > > + return sign ? min : tar_atol256_max; > > + return 0; > > +} > > + > > +inline static long tar_atol(long min) { > > + return tar_atol256(min); > > +} > > + > > +long tar_atosl() { > > + long n = tar_atol(-1); > > + if (tar_atosl_min) { > > + __errno_location(); > > + return 0; > > + } > > + if (n > 0) > > + return 0; > > + return n; > > +} > > diff --git a/gcc/tree-vect-loop-manip.cc b/gcc/tree-vect-loop-manip.cc > > index > 76d4979c0b3b374dcaacf6825a95a8714114a63b..9bacaa182a3919cae1cb99dfc > 5ae4923e1f93376 100644 > > --- a/gcc/tree-vect-loop-manip.cc > > +++ b/gcc/tree-vect-loop-manip.cc > > @@ -1719,8 +1719,6 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class loop > *loop, edge loop_exit, > > /* Now link the alternative exits. */ > > if (multiple_exits_p) > > { > > - set_immediate_dominator (CDI_DOMINATORS, new_preheader, > > - main_loop_exit_block); > > for (auto gsi_from = gsi_start_phis (loop->header), > > gsi_to = gsi_start_phis (new_preheader); > > !gsi_end_p (gsi_from) && !gsi_end_p (gsi_to); > > @@ -1776,7 +1774,14 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class loop > *loop, edge loop_exit, > > { > > update_loop = new_loop; > > for (edge e : get_loop_exit_edges (loop)) > > - doms.safe_push (e->dest); > > + { > > + /* Basic blocks that the old loop dominated are now dominated by > > + the new loop and so we have to update those. */ > > + for (auto bb : get_all_dominated_blocks (CDI_DOMINATORS, e->src)) > > + if (!flow_bb_inside_loop_p (loop, bb)) > > + doms.safe_push (bb); > > + doms.safe_push (e->dest); > > + } > > I think you'll get duplicate blocks that way. Maybe simplify this > all by instead doing > > auto doms = get_all_dominated_blocks (CDI_DOMINATORS, loop->header); > for (unsigned i = 0; i < doms.length (); ++i) > if (flow_bb_inside_loop_p (loop, doms[i])) > doms.unordered_remove (i); > > ? > > OK with that change, but really we should see to avoid this > quadraticness :/ It's probably not too bad right now given we have > quite some restrictions on vectorizing loops with multiple exits, > but I suggest you try an artificial testcase with the "same" > loop repeated N times to see whether dominance compute creeps up > in the profile. > > Richard.
On Tue, 9 Jan 2024, Tamar Christina wrote: > > > > -----Original Message----- > > From: Richard Biener <rguenther@suse.de> > > Sent: Tuesday, January 9, 2024 12:26 PM > > To: Tamar Christina <Tamar.Christina@arm.com> > > Cc: gcc-patches@gcc.gnu.org; nd <nd@arm.com>; jlaw@ventanamicro.com > > Subject: RE: [PATCH]middle-end: Fix dominators updates when peeling with > > multiple exits [PR113144] > > > > On Tue, 9 Jan 2024, Tamar Christina wrote: > > > > > > This makes it quadratic in the number of vectorized early exit loops > > > > in a function. The vectorizer CFG manipulation operates in a local > > > > enough bubble that programmatic updating of dominators should be > > > > possible (after all we manage to produce correct SSA form!), the > > > > proposed change gets us too far off to a point where re-computating > > > > dominance info is likely cheaper (but no, we shouldn't do this either). > > > > > > > > Can you instead give manual updating a try again? I think > > > > versioning should produce up-to-date dominator info, it's only > > > > when you redirect branches during peeling that you'd need > > > > adjustments - but IIRC we're never introducing new merges? > > > > > > > > IIRC we can't wipe dominators during transform since we query them > > > > during code generation. We possibly could code generate all > > > > CFG manipulations of all vectorized loops, recompute all dominators > > > > and then do code generation of all vectorized loops. > > > > > > > > But then we're doing a loop transform and the exits will ultimatively > > > > end up in the same place, so the CFG and dominator update is bound to > > > > where the original exits went to. > > > > > > Yeah that's a fair point, the issue is specifically with at_exit. So how about: > > > > > > When we peel at_exit we are moving the new loop at the exit of the previous > > > loop. This means that the blocks outside the loop dat the previous loop used to > > > dominate are no longer being dominated by it. > > > > Hmm, indeed. Note this does make the dominator update O(function-size) > > and when vectorizing multiple loops in a function this becomes > > quadratic. That's quite unfortunate so I wonder if we can delay the > > update to the parts we do not need up-to-date dominators during > > vectorization (of course it gets fragile with having only partly > > correct dominators). > > Fair, I created https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113290 and will > tackle it when I add SLP support in GCC 15. > > I think the problem is, and the reason we do early dominator correction and > validation is because the same function is used by loop distribution. > > But you're right that during vectorization we perform dominators update twice > now. We're performing it at least once per multi-exit loop that is vectorized, covering all downstream blocks. > So Maybe we should have a parameter to indicate whether dominators should > be updated? I think we should possibly try making loop distribution use another mechanism for its copying ... there's duplicate_loop_body_to_header_edge that's also used by loop_version, the core parts doing the new loop creation could be split out and the detail how the final CFG is set up be retained in two workers. Richard. > Thanks, > Tamar > > > > > > The new dominators however are hard to predict since if the loop has multiple > > > exits and all the exits are an "early" one then we always execute the scalar > > > loop. In this case the scalar loop can completely dominate the new loop. > > > > > > If we later have skip_vector then there's an additional skip edge added that > > > might change the dominators. > > > > > > The previous patch would force an update of all blocks reachable from the new > > > exits. This one updates *only* blocks that we know the scalar exits dominated. > > > > > > For the examples this reduces the blocks to update from 18 to 3. > > > > > > Bootstrapped Regtested on aarch64-none-linux-gnu, x86_64-pc-linux-gnu > > > and no issues normally and with --enable-checking=release --enable-lto > > > --with-build-config=bootstrap-O3 --enable-checking=yes,rtl,extra. > > > > > > Ok for master? > > > > See below. > > > > > Thanks, > > > Tamar > > > > > > gcc/ChangeLog: > > > > > > PR tree-optimization/113144 > > > PR tree-optimization/113145 > > > * tree-vect-loop-manip.cc (slpeel_tree_duplicate_loop_to_edge_cfg): > > > Update all BB that the original exits dominated. > > > > > > gcc/testsuite/ChangeLog: > > > > > > PR tree-optimization/113144 > > > PR tree-optimization/113145 > > > * gcc.dg/vect/vect-early-break_94-pr113144.c: New test. > > > > > > --- inline copy of patch --- > > > > > > diff --git a/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > > b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > > > new file mode 100644 > > > index > > 0000000000000000000000000000000000000000..903fe7be6621e81db6f294 > > 41e4309fa213d027c5 > > > --- /dev/null > > > +++ b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > > > @@ -0,0 +1,41 @@ > > > +/* { dg-do compile } */ > > > +/* { dg-add-options vect_early_break } */ > > > +/* { dg-require-effective-target vect_early_break } */ > > > +/* { dg-require-effective-target vect_int } */ > > > + > > > +/* { dg-final { scan-tree-dump "LOOP VECTORIZED" "vect" } } */ > > > + > > > +long tar_atol256_max, tar_atol256_size, tar_atosl_min; > > > +char tar_atol256_s; > > > +void __errno_location(); > > > + > > > + > > > +inline static long tar_atol256(long min) { > > > + char c; > > > + int sign; > > > + c = tar_atol256_s; > > > + sign = c; > > > + while (tar_atol256_size) { > > > + if (c != sign) > > > + return sign ? min : tar_atol256_max; > > > + c = tar_atol256_size--; > > > + } > > > + if ((c & 128) != (sign & 128)) > > > + return sign ? min : tar_atol256_max; > > > + return 0; > > > +} > > > + > > > +inline static long tar_atol(long min) { > > > + return tar_atol256(min); > > > +} > > > + > > > +long tar_atosl() { > > > + long n = tar_atol(-1); > > > + if (tar_atosl_min) { > > > + __errno_location(); > > > + return 0; > > > + } > > > + if (n > 0) > > > + return 0; > > > + return n; > > > +} > > > diff --git a/gcc/tree-vect-loop-manip.cc b/gcc/tree-vect-loop-manip.cc > > > index > > 76d4979c0b3b374dcaacf6825a95a8714114a63b..9bacaa182a3919cae1cb99dfc > > 5ae4923e1f93376 100644 > > > --- a/gcc/tree-vect-loop-manip.cc > > > +++ b/gcc/tree-vect-loop-manip.cc > > > @@ -1719,8 +1719,6 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class loop > > *loop, edge loop_exit, > > > /* Now link the alternative exits. */ > > > if (multiple_exits_p) > > > { > > > - set_immediate_dominator (CDI_DOMINATORS, new_preheader, > > > - main_loop_exit_block); > > > for (auto gsi_from = gsi_start_phis (loop->header), > > > gsi_to = gsi_start_phis (new_preheader); > > > !gsi_end_p (gsi_from) && !gsi_end_p (gsi_to); > > > @@ -1776,7 +1774,14 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class loop > > *loop, edge loop_exit, > > > { > > > update_loop = new_loop; > > > for (edge e : get_loop_exit_edges (loop)) > > > - doms.safe_push (e->dest); > > > + { > > > + /* Basic blocks that the old loop dominated are now dominated by > > > + the new loop and so we have to update those. */ > > > + for (auto bb : get_all_dominated_blocks (CDI_DOMINATORS, e->src)) > > > + if (!flow_bb_inside_loop_p (loop, bb)) > > > + doms.safe_push (bb); > > > + doms.safe_push (e->dest); > > > + } > > > > I think you'll get duplicate blocks that way. Maybe simplify this > > all by instead doing > > > > auto doms = get_all_dominated_blocks (CDI_DOMINATORS, loop->header); > > for (unsigned i = 0; i < doms.length (); ++i) > > if (flow_bb_inside_loop_p (loop, doms[i])) > > doms.unordered_remove (i); > > > > ? > > > > OK with that change, but really we should see to avoid this > > quadraticness :/ It's probably not too bad right now given we have > > quite some restrictions on vectorizing loops with multiple exits, > > but I suggest you try an artificial testcase with the "same" > > loop repeated N times to see whether dominance compute creeps up > > in the profile. > > > > Richard. >
On Tue, 9 Jan 2024, Richard Biener wrote: > On Tue, 9 Jan 2024, Tamar Christina wrote: > > > > > > > > -----Original Message----- > > > From: Richard Biener <rguenther@suse.de> > > > Sent: Tuesday, January 9, 2024 12:26 PM > > > To: Tamar Christina <Tamar.Christina@arm.com> > > > Cc: gcc-patches@gcc.gnu.org; nd <nd@arm.com>; jlaw@ventanamicro.com > > > Subject: RE: [PATCH]middle-end: Fix dominators updates when peeling with > > > multiple exits [PR113144] > > > > > > On Tue, 9 Jan 2024, Tamar Christina wrote: > > > > > > > > This makes it quadratic in the number of vectorized early exit loops > > > > > in a function. The vectorizer CFG manipulation operates in a local > > > > > enough bubble that programmatic updating of dominators should be > > > > > possible (after all we manage to produce correct SSA form!), the > > > > > proposed change gets us too far off to a point where re-computating > > > > > dominance info is likely cheaper (but no, we shouldn't do this either). > > > > > > > > > > Can you instead give manual updating a try again? I think > > > > > versioning should produce up-to-date dominator info, it's only > > > > > when you redirect branches during peeling that you'd need > > > > > adjustments - but IIRC we're never introducing new merges? > > > > > > > > > > IIRC we can't wipe dominators during transform since we query them > > > > > during code generation. We possibly could code generate all > > > > > CFG manipulations of all vectorized loops, recompute all dominators > > > > > and then do code generation of all vectorized loops. > > > > > > > > > > But then we're doing a loop transform and the exits will ultimatively > > > > > end up in the same place, so the CFG and dominator update is bound to > > > > > where the original exits went to. > > > > > > > > Yeah that's a fair point, the issue is specifically with at_exit. So how about: > > > > > > > > When we peel at_exit we are moving the new loop at the exit of the previous > > > > loop. This means that the blocks outside the loop dat the previous loop used to > > > > dominate are no longer being dominated by it. > > > > > > Hmm, indeed. Note this does make the dominator update O(function-size) > > > and when vectorizing multiple loops in a function this becomes > > > quadratic. That's quite unfortunate so I wonder if we can delay the > > > update to the parts we do not need up-to-date dominators during > > > vectorization (of course it gets fragile with having only partly > > > correct dominators). > > > > Fair, I created https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113290 and will > > tackle it when I add SLP support in GCC 15. > > > > I think the problem is, and the reason we do early dominator correction and > > validation is because the same function is used by loop distribution. > > > > But you're right that during vectorization we perform dominators update twice > > now. > > We're performing it at least once per multi-exit loop that is vectorized, > covering all downstream blocks. That is, consider sth like int a[77]; int bar (); void foo () { int val; #define LOOP \ val = bar (); \ for (int i = 0; i < 77; ++i) \ { \ if (a[i] == val) \ break; \ a[i]++; \ } #define LOOP10 LOOP LOOP LOOP LOOP LOOP LOOP LOOP LOOP LOOP LOOP #define LOOP100 LOOP10 LOOP10 LOOP10 LOOP10 LOOP10 LOOP10 LOOP10 LOOP10 LOOP10 LOOP10 #define LOOP1000 LOOP100 LOOP100 LOOP100 LOOP100 LOOP100 LOOP100 LOOP100 LOOP100 LOOP100 LOOP100 LOOP1000 } where on x86_64 with -O3 -msse4.1 -fno-gcse -fno-gcse-after-reload we're currently "fine" (calling iterate_fix_dominators 2000 times). But with geting all dominated blocks you should get every block to exit "fixed" and maybe get dominance compute to show up in the profile. Richard. > > So Maybe we should have a parameter to indicate whether dominators should > > be updated? > > I think we should possibly try making loop distribution use another > mechanism for its copying ... there's duplicate_loop_body_to_header_edge > that's also used by loop_version, the core parts doing the new > loop creation could be split out and the detail how the final CFG > is set up be retained in two workers. > > Richard. > > > Thanks, > > Tamar > > > > > > > > > The new dominators however are hard to predict since if the loop has multiple > > > > exits and all the exits are an "early" one then we always execute the scalar > > > > loop. In this case the scalar loop can completely dominate the new loop. > > > > > > > > If we later have skip_vector then there's an additional skip edge added that > > > > might change the dominators. > > > > > > > > The previous patch would force an update of all blocks reachable from the new > > > > exits. This one updates *only* blocks that we know the scalar exits dominated. > > > > > > > > For the examples this reduces the blocks to update from 18 to 3. > > > > > > > > Bootstrapped Regtested on aarch64-none-linux-gnu, x86_64-pc-linux-gnu > > > > and no issues normally and with --enable-checking=release --enable-lto > > > > --with-build-config=bootstrap-O3 --enable-checking=yes,rtl,extra. > > > > > > > > Ok for master? > > > > > > See below. > > > > > > > Thanks, > > > > Tamar > > > > > > > > gcc/ChangeLog: > > > > > > > > PR tree-optimization/113144 > > > > PR tree-optimization/113145 > > > > * tree-vect-loop-manip.cc (slpeel_tree_duplicate_loop_to_edge_cfg): > > > > Update all BB that the original exits dominated. > > > > > > > > gcc/testsuite/ChangeLog: > > > > > > > > PR tree-optimization/113144 > > > > PR tree-optimization/113145 > > > > * gcc.dg/vect/vect-early-break_94-pr113144.c: New test. > > > > > > > > --- inline copy of patch --- > > > > > > > > diff --git a/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > > > b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > > > > new file mode 100644 > > > > index > > > 0000000000000000000000000000000000000000..903fe7be6621e81db6f294 > > > 41e4309fa213d027c5 > > > > --- /dev/null > > > > +++ b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > > > > @@ -0,0 +1,41 @@ > > > > +/* { dg-do compile } */ > > > > +/* { dg-add-options vect_early_break } */ > > > > +/* { dg-require-effective-target vect_early_break } */ > > > > +/* { dg-require-effective-target vect_int } */ > > > > + > > > > +/* { dg-final { scan-tree-dump "LOOP VECTORIZED" "vect" } } */ > > > > + > > > > +long tar_atol256_max, tar_atol256_size, tar_atosl_min; > > > > +char tar_atol256_s; > > > > +void __errno_location(); > > > > + > > > > + > > > > +inline static long tar_atol256(long min) { > > > > + char c; > > > > + int sign; > > > > + c = tar_atol256_s; > > > > + sign = c; > > > > + while (tar_atol256_size) { > > > > + if (c != sign) > > > > + return sign ? min : tar_atol256_max; > > > > + c = tar_atol256_size--; > > > > + } > > > > + if ((c & 128) != (sign & 128)) > > > > + return sign ? min : tar_atol256_max; > > > > + return 0; > > > > +} > > > > + > > > > +inline static long tar_atol(long min) { > > > > + return tar_atol256(min); > > > > +} > > > > + > > > > +long tar_atosl() { > > > > + long n = tar_atol(-1); > > > > + if (tar_atosl_min) { > > > > + __errno_location(); > > > > + return 0; > > > > + } > > > > + if (n > 0) > > > > + return 0; > > > > + return n; > > > > +} > > > > diff --git a/gcc/tree-vect-loop-manip.cc b/gcc/tree-vect-loop-manip.cc > > > > index > > > 76d4979c0b3b374dcaacf6825a95a8714114a63b..9bacaa182a3919cae1cb99dfc > > > 5ae4923e1f93376 100644 > > > > --- a/gcc/tree-vect-loop-manip.cc > > > > +++ b/gcc/tree-vect-loop-manip.cc > > > > @@ -1719,8 +1719,6 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class loop > > > *loop, edge loop_exit, > > > > /* Now link the alternative exits. */ > > > > if (multiple_exits_p) > > > > { > > > > - set_immediate_dominator (CDI_DOMINATORS, new_preheader, > > > > - main_loop_exit_block); > > > > for (auto gsi_from = gsi_start_phis (loop->header), > > > > gsi_to = gsi_start_phis (new_preheader); > > > > !gsi_end_p (gsi_from) && !gsi_end_p (gsi_to); > > > > @@ -1776,7 +1774,14 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class loop > > > *loop, edge loop_exit, > > > > { > > > > update_loop = new_loop; > > > > for (edge e : get_loop_exit_edges (loop)) > > > > - doms.safe_push (e->dest); > > > > + { > > > > + /* Basic blocks that the old loop dominated are now dominated by > > > > + the new loop and so we have to update those. */ > > > > + for (auto bb : get_all_dominated_blocks (CDI_DOMINATORS, e->src)) > > > > + if (!flow_bb_inside_loop_p (loop, bb)) > > > > + doms.safe_push (bb); > > > > + doms.safe_push (e->dest); > > > > + } > > > > > > I think you'll get duplicate blocks that way. Maybe simplify this > > > all by instead doing > > > > > > auto doms = get_all_dominated_blocks (CDI_DOMINATORS, loop->header); > > > for (unsigned i = 0; i < doms.length (); ++i) > > > if (flow_bb_inside_loop_p (loop, doms[i])) > > > doms.unordered_remove (i); > > > > > > ? > > > > > > OK with that change, but really we should see to avoid this > > > quadraticness :/ It's probably not too bad right now given we have > > > quite some restrictions on vectorizing loops with multiple exits, > > > but I suggest you try an artificial testcase with the "same" > > > loop repeated N times to see whether dominance compute creeps up > > > in the profile. > > > > > > Richard. > > > >
> -----Original Message----- > From: Richard Biener <rguenther@suse.de> > Sent: Tuesday, January 9, 2024 1:51 PM > To: Tamar Christina <Tamar.Christina@arm.com> > Cc: gcc-patches@gcc.gnu.org; nd <nd@arm.com>; jlaw@ventanamicro.com > Subject: RE: [PATCH]middle-end: Fix dominators updates when peeling with > multiple exits [PR113144] > > On Tue, 9 Jan 2024, Richard Biener wrote: > > > On Tue, 9 Jan 2024, Tamar Christina wrote: > > > > > > > > > > > > -----Original Message----- > > > > From: Richard Biener <rguenther@suse.de> > > > > Sent: Tuesday, January 9, 2024 12:26 PM > > > > To: Tamar Christina <Tamar.Christina@arm.com> > > > > Cc: gcc-patches@gcc.gnu.org; nd <nd@arm.com>; jlaw@ventanamicro.com > > > > Subject: RE: [PATCH]middle-end: Fix dominators updates when peeling with > > > > multiple exits [PR113144] > > > > > > > > On Tue, 9 Jan 2024, Tamar Christina wrote: > > > > > > > > > > This makes it quadratic in the number of vectorized early exit loops > > > > > > in a function. The vectorizer CFG manipulation operates in a local > > > > > > enough bubble that programmatic updating of dominators should be > > > > > > possible (after all we manage to produce correct SSA form!), the > > > > > > proposed change gets us too far off to a point where re-computating > > > > > > dominance info is likely cheaper (but no, we shouldn't do this either). > > > > > > > > > > > > Can you instead give manual updating a try again? I think > > > > > > versioning should produce up-to-date dominator info, it's only > > > > > > when you redirect branches during peeling that you'd need > > > > > > adjustments - but IIRC we're never introducing new merges? > > > > > > > > > > > > IIRC we can't wipe dominators during transform since we query them > > > > > > during code generation. We possibly could code generate all > > > > > > CFG manipulations of all vectorized loops, recompute all dominators > > > > > > and then do code generation of all vectorized loops. > > > > > > > > > > > > But then we're doing a loop transform and the exits will ultimatively > > > > > > end up in the same place, so the CFG and dominator update is bound to > > > > > > where the original exits went to. > > > > > > > > > > Yeah that's a fair point, the issue is specifically with at_exit. So how about: > > > > > > > > > > When we peel at_exit we are moving the new loop at the exit of the > previous > > > > > loop. This means that the blocks outside the loop dat the previous loop > used to > > > > > dominate are no longer being dominated by it. > > > > > > > > Hmm, indeed. Note this does make the dominator update O(function-size) > > > > and when vectorizing multiple loops in a function this becomes > > > > quadratic. That's quite unfortunate so I wonder if we can delay the > > > > update to the parts we do not need up-to-date dominators during > > > > vectorization (of course it gets fragile with having only partly > > > > correct dominators). > > > > > > Fair, I created https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113290 and will > > > tackle it when I add SLP support in GCC 15. > > > > > > I think the problem is, and the reason we do early dominator correction and > > > validation is because the same function is used by loop distribution. > > > > > > But you're right that during vectorization we perform dominators update twice > > > now. > > > > We're performing it at least once per multi-exit loop that is vectorized, > > covering all downstream blocks. > > That is, consider sth like > > int a[77]; > > int bar (); > void foo () > { > int val; > #define LOOP \ > val = bar (); \ > for (int i = 0; i < 77; ++i) \ > { \ > if (a[i] == val) \ > break; \ > a[i]++; \ > } > #define LOOP10 LOOP LOOP LOOP LOOP LOOP LOOP LOOP LOOP LOOP LOOP > #define LOOP100 LOOP10 LOOP10 LOOP10 LOOP10 LOOP10 LOOP10 LOOP10 > LOOP10 > LOOP10 LOOP10 > #define LOOP1000 LOOP100 LOOP100 LOOP100 LOOP100 LOOP100 LOOP100 > LOOP100 > LOOP100 LOOP100 LOOP100 > LOOP1000 > } > > where on x86_64 with -O3 -msse4.1 -fno-gcse -fno-gcse-after-reload we're > currently "fine" (calling iterate_fix_dominators 2000 times). But > with geting all dominated blocks you should get every block to exit > "fixed" and maybe get dominance compute to show up in the profile. > Yeah, that makes sense. If we can move loop distribution to a different method, then we can just perform dominators update only once for all loops at the end of vectorization right? Thanks, Tamar > Richard. > > > > > So Maybe we should have a parameter to indicate whether dominators should > > > be updated? > > > > I think we should possibly try making loop distribution use another > > mechanism for its copying ... there's duplicate_loop_body_to_header_edge > > that's also used by loop_version, the core parts doing the new > > loop creation could be split out and the detail how the final CFG > > is set up be retained in two workers. > > > > Richard. > > > > > Thanks, > > > Tamar > > > > > > > > > > > > The new dominators however are hard to predict since if the loop has > multiple > > > > > exits and all the exits are an "early" one then we always execute the scalar > > > > > loop. In this case the scalar loop can completely dominate the new loop. > > > > > > > > > > If we later have skip_vector then there's an additional skip edge added that > > > > > might change the dominators. > > > > > > > > > > The previous patch would force an update of all blocks reachable from the > new > > > > > exits. This one updates *only* blocks that we know the scalar exits > dominated. > > > > > > > > > > For the examples this reduces the blocks to update from 18 to 3. > > > > > > > > > > Bootstrapped Regtested on aarch64-none-linux-gnu, x86_64-pc-linux-gnu > > > > > and no issues normally and with --enable-checking=release --enable-lto > > > > > --with-build-config=bootstrap-O3 --enable-checking=yes,rtl,extra. > > > > > > > > > > Ok for master? > > > > > > > > See below. > > > > > > > > > Thanks, > > > > > Tamar > > > > > > > > > > gcc/ChangeLog: > > > > > > > > > > PR tree-optimization/113144 > > > > > PR tree-optimization/113145 > > > > > * tree-vect-loop-manip.cc (slpeel_tree_duplicate_loop_to_edge_cfg): > > > > > Update all BB that the original exits dominated. > > > > > > > > > > gcc/testsuite/ChangeLog: > > > > > > > > > > PR tree-optimization/113144 > > > > > PR tree-optimization/113145 > > > > > * gcc.dg/vect/vect-early-break_94-pr113144.c: New test. > > > > > > > > > > --- inline copy of patch --- > > > > > > > > > > diff --git a/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > > > > b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > > > > > new file mode 100644 > > > > > index > > > > > 0000000000000000000000000000000000000000..903fe7be6621e81db6f294 > > > > 41e4309fa213d027c5 > > > > > --- /dev/null > > > > > +++ b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > > > > > @@ -0,0 +1,41 @@ > > > > > +/* { dg-do compile } */ > > > > > +/* { dg-add-options vect_early_break } */ > > > > > +/* { dg-require-effective-target vect_early_break } */ > > > > > +/* { dg-require-effective-target vect_int } */ > > > > > + > > > > > +/* { dg-final { scan-tree-dump "LOOP VECTORIZED" "vect" } } */ > > > > > + > > > > > +long tar_atol256_max, tar_atol256_size, tar_atosl_min; > > > > > +char tar_atol256_s; > > > > > +void __errno_location(); > > > > > + > > > > > + > > > > > +inline static long tar_atol256(long min) { > > > > > + char c; > > > > > + int sign; > > > > > + c = tar_atol256_s; > > > > > + sign = c; > > > > > + while (tar_atol256_size) { > > > > > + if (c != sign) > > > > > + return sign ? min : tar_atol256_max; > > > > > + c = tar_atol256_size--; > > > > > + } > > > > > + if ((c & 128) != (sign & 128)) > > > > > + return sign ? min : tar_atol256_max; > > > > > + return 0; > > > > > +} > > > > > + > > > > > +inline static long tar_atol(long min) { > > > > > + return tar_atol256(min); > > > > > +} > > > > > + > > > > > +long tar_atosl() { > > > > > + long n = tar_atol(-1); > > > > > + if (tar_atosl_min) { > > > > > + __errno_location(); > > > > > + return 0; > > > > > + } > > > > > + if (n > 0) > > > > > + return 0; > > > > > + return n; > > > > > +} > > > > > diff --git a/gcc/tree-vect-loop-manip.cc b/gcc/tree-vect-loop-manip.cc > > > > > index > > > > > 76d4979c0b3b374dcaacf6825a95a8714114a63b..9bacaa182a3919cae1cb99dfc > > > > 5ae4923e1f93376 100644 > > > > > --- a/gcc/tree-vect-loop-manip.cc > > > > > +++ b/gcc/tree-vect-loop-manip.cc > > > > > @@ -1719,8 +1719,6 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class > loop > > > > *loop, edge loop_exit, > > > > > /* Now link the alternative exits. */ > > > > > if (multiple_exits_p) > > > > > { > > > > > - set_immediate_dominator (CDI_DOMINATORS, new_preheader, > > > > > - main_loop_exit_block); > > > > > for (auto gsi_from = gsi_start_phis (loop->header), > > > > > gsi_to = gsi_start_phis (new_preheader); > > > > > !gsi_end_p (gsi_from) && !gsi_end_p (gsi_to); > > > > > @@ -1776,7 +1774,14 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class > loop > > > > *loop, edge loop_exit, > > > > > { > > > > > update_loop = new_loop; > > > > > for (edge e : get_loop_exit_edges (loop)) > > > > > - doms.safe_push (e->dest); > > > > > + { > > > > > + /* Basic blocks that the old loop dominated are now dominated > by > > > > > + the new loop and so we have to update those. */ > > > > > + for (auto bb : get_all_dominated_blocks (CDI_DOMINATORS, e- > >src)) > > > > > + if (!flow_bb_inside_loop_p (loop, bb)) > > > > > + doms.safe_push (bb); > > > > > + doms.safe_push (e->dest); > > > > > + } > > > > > > > > I think you'll get duplicate blocks that way. Maybe simplify this > > > > all by instead doing > > > > > > > > auto doms = get_all_dominated_blocks (CDI_DOMINATORS, loop- > >header); > > > > for (unsigned i = 0; i < doms.length (); ++i) > > > > if (flow_bb_inside_loop_p (loop, doms[i])) > > > > doms.unordered_remove (i); > > > > > > > > ? > > > > > > > > OK with that change, but really we should see to avoid this > > > > quadraticness :/ It's probably not too bad right now given we have > > > > quite some restrictions on vectorizing loops with multiple exits, > > > > but I suggest you try an artificial testcase with the "same" > > > > loop repeated N times to see whether dominance compute creeps up > > > > in the profile. > > > > > > > > Richard. > > > > > > > > > -- > Richard Biener <rguenther@suse.de> > SUSE Software Solutions Germany GmbH, > Frankenstrasse 146, 90461 Nuernberg, Germany; > GF: Ivo Totev, Andrew McDonald, Werner Knoblich; (HRB 36809, AG Nuernberg)
On Tue, 9 Jan 2024, Tamar Christina wrote: > > -----Original Message----- > > From: Richard Biener <rguenther@suse.de> > > Sent: Tuesday, January 9, 2024 1:51 PM > > To: Tamar Christina <Tamar.Christina@arm.com> > > Cc: gcc-patches@gcc.gnu.org; nd <nd@arm.com>; jlaw@ventanamicro.com > > Subject: RE: [PATCH]middle-end: Fix dominators updates when peeling with > > multiple exits [PR113144] > > > > On Tue, 9 Jan 2024, Richard Biener wrote: > > > > > On Tue, 9 Jan 2024, Tamar Christina wrote: > > > > > > > > > > > > > > > > -----Original Message----- > > > > > From: Richard Biener <rguenther@suse.de> > > > > > Sent: Tuesday, January 9, 2024 12:26 PM > > > > > To: Tamar Christina <Tamar.Christina@arm.com> > > > > > Cc: gcc-patches@gcc.gnu.org; nd <nd@arm.com>; jlaw@ventanamicro.com > > > > > Subject: RE: [PATCH]middle-end: Fix dominators updates when peeling with > > > > > multiple exits [PR113144] > > > > > > > > > > On Tue, 9 Jan 2024, Tamar Christina wrote: > > > > > > > > > > > > This makes it quadratic in the number of vectorized early exit loops > > > > > > > in a function. The vectorizer CFG manipulation operates in a local > > > > > > > enough bubble that programmatic updating of dominators should be > > > > > > > possible (after all we manage to produce correct SSA form!), the > > > > > > > proposed change gets us too far off to a point where re-computating > > > > > > > dominance info is likely cheaper (but no, we shouldn't do this either). > > > > > > > > > > > > > > Can you instead give manual updating a try again? I think > > > > > > > versioning should produce up-to-date dominator info, it's only > > > > > > > when you redirect branches during peeling that you'd need > > > > > > > adjustments - but IIRC we're never introducing new merges? > > > > > > > > > > > > > > IIRC we can't wipe dominators during transform since we query them > > > > > > > during code generation. We possibly could code generate all > > > > > > > CFG manipulations of all vectorized loops, recompute all dominators > > > > > > > and then do code generation of all vectorized loops. > > > > > > > > > > > > > > But then we're doing a loop transform and the exits will ultimatively > > > > > > > end up in the same place, so the CFG and dominator update is bound to > > > > > > > where the original exits went to. > > > > > > > > > > > > Yeah that's a fair point, the issue is specifically with at_exit. So how about: > > > > > > > > > > > > When we peel at_exit we are moving the new loop at the exit of the > > previous > > > > > > loop. This means that the blocks outside the loop dat the previous loop > > used to > > > > > > dominate are no longer being dominated by it. > > > > > > > > > > Hmm, indeed. Note this does make the dominator update O(function-size) > > > > > and when vectorizing multiple loops in a function this becomes > > > > > quadratic. That's quite unfortunate so I wonder if we can delay the > > > > > update to the parts we do not need up-to-date dominators during > > > > > vectorization (of course it gets fragile with having only partly > > > > > correct dominators). > > > > > > > > Fair, I created https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113290 and will > > > > tackle it when I add SLP support in GCC 15. > > > > > > > > I think the problem is, and the reason we do early dominator correction and > > > > validation is because the same function is used by loop distribution. > > > > > > > > But you're right that during vectorization we perform dominators update twice > > > > now. > > > > > > We're performing it at least once per multi-exit loop that is vectorized, > > > covering all downstream blocks. > > > > That is, consider sth like > > > > int a[77]; > > > > int bar (); > > void foo () > > { > > int val; > > #define LOOP \ > > val = bar (); \ > > for (int i = 0; i < 77; ++i) \ > > { \ > > if (a[i] == val) \ > > break; \ > > a[i]++; \ > > } > > #define LOOP10 LOOP LOOP LOOP LOOP LOOP LOOP LOOP LOOP LOOP LOOP > > #define LOOP100 LOOP10 LOOP10 LOOP10 LOOP10 LOOP10 LOOP10 LOOP10 > > LOOP10 > > LOOP10 LOOP10 > > #define LOOP1000 LOOP100 LOOP100 LOOP100 LOOP100 LOOP100 LOOP100 > > LOOP100 > > LOOP100 LOOP100 LOOP100 > > LOOP1000 > > } > > > > where on x86_64 with -O3 -msse4.1 -fno-gcse -fno-gcse-after-reload we're > > currently "fine" (calling iterate_fix_dominators 2000 times). But > > with geting all dominated blocks you should get every block to exit > > "fixed" and maybe get dominance compute to show up in the profile. > > > > Yeah, that makes sense. If we can move loop distribution to a different method, > then we can just perform dominators update only once for all loops at the end > of vectorization right? Maybe. We'll have to see. It would be good to have a way to mark dominator regions as invalid (you could simply disconnect the nodes from the tree). What we know is that within loops dominator queries should be OK. I still think the invalidated region is bound, for example to blocks dominating the sons of the scalar loop header, that is, I doubt you really need to use get_all_dominated_blocks, get_dominated_by should eventually suffice (or get_dominated_by_region with the region being the scalar loop body and maybe exit blocks)? > Thanks, > Tamar > > > Richard. > > > > > > > > So Maybe we should have a parameter to indicate whether dominators should > > > > be updated? > > > > > > I think we should possibly try making loop distribution use another > > > mechanism for its copying ... there's duplicate_loop_body_to_header_edge > > > that's also used by loop_version, the core parts doing the new > > > loop creation could be split out and the detail how the final CFG > > > is set up be retained in two workers. > > > > > > Richard. > > > > > > > Thanks, > > > > Tamar > > > > > > > > > > > > > > > The new dominators however are hard to predict since if the loop has > > multiple > > > > > > exits and all the exits are an "early" one then we always execute the scalar > > > > > > loop. In this case the scalar loop can completely dominate the new loop. > > > > > > > > > > > > If we later have skip_vector then there's an additional skip edge added that > > > > > > might change the dominators. > > > > > > > > > > > > The previous patch would force an update of all blocks reachable from the > > new > > > > > > exits. This one updates *only* blocks that we know the scalar exits > > dominated. > > > > > > > > > > > > For the examples this reduces the blocks to update from 18 to 3. > > > > > > > > > > > > Bootstrapped Regtested on aarch64-none-linux-gnu, x86_64-pc-linux-gnu > > > > > > and no issues normally and with --enable-checking=release --enable-lto > > > > > > --with-build-config=bootstrap-O3 --enable-checking=yes,rtl,extra. > > > > > > > > > > > > Ok for master? > > > > > > > > > > See below. > > > > > > > > > > > Thanks, > > > > > > Tamar > > > > > > > > > > > > gcc/ChangeLog: > > > > > > > > > > > > PR tree-optimization/113144 > > > > > > PR tree-optimization/113145 > > > > > > * tree-vect-loop-manip.cc (slpeel_tree_duplicate_loop_to_edge_cfg): > > > > > > Update all BB that the original exits dominated. > > > > > > > > > > > > gcc/testsuite/ChangeLog: > > > > > > > > > > > > PR tree-optimization/113144 > > > > > > PR tree-optimization/113145 > > > > > > * gcc.dg/vect/vect-early-break_94-pr113144.c: New test. > > > > > > > > > > > > --- inline copy of patch --- > > > > > > > > > > > > diff --git a/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > > > > > b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > > > > > > new file mode 100644 > > > > > > index > > > > > > > 0000000000000000000000000000000000000000..903fe7be6621e81db6f294 > > > > > 41e4309fa213d027c5 > > > > > > --- /dev/null > > > > > > +++ b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c > > > > > > @@ -0,0 +1,41 @@ > > > > > > +/* { dg-do compile } */ > > > > > > +/* { dg-add-options vect_early_break } */ > > > > > > +/* { dg-require-effective-target vect_early_break } */ > > > > > > +/* { dg-require-effective-target vect_int } */ > > > > > > + > > > > > > +/* { dg-final { scan-tree-dump "LOOP VECTORIZED" "vect" } } */ > > > > > > + > > > > > > +long tar_atol256_max, tar_atol256_size, tar_atosl_min; > > > > > > +char tar_atol256_s; > > > > > > +void __errno_location(); > > > > > > + > > > > > > + > > > > > > +inline static long tar_atol256(long min) { > > > > > > + char c; > > > > > > + int sign; > > > > > > + c = tar_atol256_s; > > > > > > + sign = c; > > > > > > + while (tar_atol256_size) { > > > > > > + if (c != sign) > > > > > > + return sign ? min : tar_atol256_max; > > > > > > + c = tar_atol256_size--; > > > > > > + } > > > > > > + if ((c & 128) != (sign & 128)) > > > > > > + return sign ? min : tar_atol256_max; > > > > > > + return 0; > > > > > > +} > > > > > > + > > > > > > +inline static long tar_atol(long min) { > > > > > > + return tar_atol256(min); > > > > > > +} > > > > > > + > > > > > > +long tar_atosl() { > > > > > > + long n = tar_atol(-1); > > > > > > + if (tar_atosl_min) { > > > > > > + __errno_location(); > > > > > > + return 0; > > > > > > + } > > > > > > + if (n > 0) > > > > > > + return 0; > > > > > > + return n; > > > > > > +} > > > > > > diff --git a/gcc/tree-vect-loop-manip.cc b/gcc/tree-vect-loop-manip.cc > > > > > > index > > > > > > > 76d4979c0b3b374dcaacf6825a95a8714114a63b..9bacaa182a3919cae1cb99dfc > > > > > 5ae4923e1f93376 100644 > > > > > > --- a/gcc/tree-vect-loop-manip.cc > > > > > > +++ b/gcc/tree-vect-loop-manip.cc > > > > > > @@ -1719,8 +1719,6 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class > > loop > > > > > *loop, edge loop_exit, > > > > > > /* Now link the alternative exits. */ > > > > > > if (multiple_exits_p) > > > > > > { > > > > > > - set_immediate_dominator (CDI_DOMINATORS, new_preheader, > > > > > > - main_loop_exit_block); > > > > > > for (auto gsi_from = gsi_start_phis (loop->header), > > > > > > gsi_to = gsi_start_phis (new_preheader); > > > > > > !gsi_end_p (gsi_from) && !gsi_end_p (gsi_to); > > > > > > @@ -1776,7 +1774,14 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class > > loop > > > > > *loop, edge loop_exit, > > > > > > { > > > > > > update_loop = new_loop; > > > > > > for (edge e : get_loop_exit_edges (loop)) > > > > > > - doms.safe_push (e->dest); > > > > > > + { > > > > > > + /* Basic blocks that the old loop dominated are now dominated > > by > > > > > > + the new loop and so we have to update those. */ > > > > > > + for (auto bb : get_all_dominated_blocks (CDI_DOMINATORS, e- > > >src)) > > > > > > + if (!flow_bb_inside_loop_p (loop, bb)) > > > > > > + doms.safe_push (bb); > > > > > > + doms.safe_push (e->dest); > > > > > > + } > > > > > > > > > > I think you'll get duplicate blocks that way. Maybe simplify this > > > > > all by instead doing > > > > > > > > > > auto doms = get_all_dominated_blocks (CDI_DOMINATORS, loop- > > >header); > > > > > for (unsigned i = 0; i < doms.length (); ++i) > > > > > if (flow_bb_inside_loop_p (loop, doms[i])) > > > > > doms.unordered_remove (i); > > > > > > > > > > ? > > > > > > > > > > OK with that change, but really we should see to avoid this > > > > > quadraticness :/ It's probably not too bad right now given we have > > > > > quite some restrictions on vectorizing loops with multiple exits, > > > > > but I suggest you try an artificial testcase with the "same" > > > > > loop repeated N times to see whether dominance compute creeps up > > > > > in the profile. > > > > > > > > > > Richard. > > > > > > > > > > > > > > -- > > Richard Biener <rguenther@suse.de> > > SUSE Software Solutions Germany GmbH, > > Frankenstrasse 146, 90461 Nuernberg, Germany; > > GF: Ivo Totev, Andrew McDonald, Werner Knoblich; (HRB 36809, AG Nuernberg) >
--- /dev/null +++ b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c @@ -0,0 +1,41 @@ +/* { dg-do compile } */ +/* { dg-add-options vect_early_break } */ +/* { dg-require-effective-target vect_early_break } */ +/* { dg-require-effective-target vect_int } */ + +/* { dg-final { scan-tree-dump "LOOP VECTORIZED" "vect" } } */ + +long tar_atol256_max, tar_atol256_size, tar_atosl_min; +char tar_atol256_s; +void __errno_location(); + + +inline static long tar_atol256(long min) { + char c; + int sign; + c = tar_atol256_s; + sign = c; + while (tar_atol256_size) { + if (c != sign) + return sign ? min : tar_atol256_max; + c = tar_atol256_size--; + } + if ((c & 128) != (sign & 128)) + return sign ? min : tar_atol256_max; + return 0; +} + +inline static long tar_atol(long min) { + return tar_atol256(min); +} + +long tar_atosl() { + long n = tar_atol(-1); + if (tar_atosl_min) { + __errno_location(); + return 0; + } + if (n > 0) + return 0; + return n; +} diff --git a/gcc/tree-vect-loop-manip.cc b/gcc/tree-vect-loop-manip.cc index 1066ea17c5674e03412b3dcd8a62ddf4dd54cf31..3810983a80c8b989be9fd9a9993642069fd39b99 100644 --- a/gcc/tree-vect-loop-manip.cc +++ b/gcc/tree-vect-loop-manip.cc @@ -1716,8 +1716,6 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class loop *loop, edge loop_exit, /* Now link the alternative exits. */ if (multiple_exits_p) { - set_immediate_dominator (CDI_DOMINATORS, new_preheader, - main_loop_exit_block); for (auto gsi_from = gsi_start_phis (loop->header), gsi_to = gsi_start_phis (new_preheader); !gsi_end_p (gsi_from) && !gsi_end_p (gsi_to); @@ -1751,12 +1749,26 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class loop *loop, edge loop_exit, /* Finally after wiring the new epilogue we need to update its main exit to the original function exit we recorded. Other exits are already - correct. */ + correct. Because of versioning, skip vectors and others we must update + the dominators of every node reachable by the new exits. */ if (multiple_exits_p) { update_loop = new_loop; - for (edge e : get_loop_exit_edges (loop)) - doms.safe_push (e->dest); + hash_set <basic_block> visited; + auto_vec <edge> workset; + edge ev; + edge_iterator ei; + workset.safe_splice (get_loop_exit_edges (loop)); + while (!workset.is_empty ()) + { + auto bb = workset.pop ()->dest; + if (visited.add (bb)) + continue; + doms.safe_push (bb); + FOR_EACH_EDGE (ev, ei, bb->succs) + workset.safe_push (ev); + } + visited.empty (); doms.safe_push (exit_dest); /* Likely a fall-through edge, so update if needed. */
Hi All, Only trying to update certain dominators doesn't seem to work very well because as the loop gets versioned, peeled, or skip_vector then we end up with very complicated control flow. This means that the final merge blocks for the loop exit are not easy to find or update. Instead of trying to pick which exits to update, this changes it to update all the blocks reachable by the new exits. This is because they'll contain common blocks with e.g. the versioned loop. It's these blocks that need an update most of the time. Bootstrapped Regtested on aarch64-none-linux-gnu and no issues. Ok for master? Thanks, Tamar gcc/ChangeLog: PR middle-end/113144 * tree-vect-loop-manip.cc (slpeel_tree_duplicate_loop_to_edge_cfg): Update all dominators reachable from exit. gcc/testsuite/ChangeLog: PR middle-end/113144 * gcc.dg/vect/vect-early-break_94-pr113144.c: New test. --- inline copy of patch -- diff --git a/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c new file mode 100644 index 0000000000000000000000000000000000000000..903fe7be6621e81db6f29441e4309fa213d027c5 -- diff --git a/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c new file mode 100644 index 0000000000000000000000000000000000000000..903fe7be6621e81db6f29441e4309fa213d027c5 --- /dev/null +++ b/gcc/testsuite/gcc.dg/vect/vect-early-break_94-pr113144.c @@ -0,0 +1,41 @@ +/* { dg-do compile } */ +/* { dg-add-options vect_early_break } */ +/* { dg-require-effective-target vect_early_break } */ +/* { dg-require-effective-target vect_int } */ + +/* { dg-final { scan-tree-dump "LOOP VECTORIZED" "vect" } } */ + +long tar_atol256_max, tar_atol256_size, tar_atosl_min; +char tar_atol256_s; +void __errno_location(); + + +inline static long tar_atol256(long min) { + char c; + int sign; + c = tar_atol256_s; + sign = c; + while (tar_atol256_size) { + if (c != sign) + return sign ? min : tar_atol256_max; + c = tar_atol256_size--; + } + if ((c & 128) != (sign & 128)) + return sign ? min : tar_atol256_max; + return 0; +} + +inline static long tar_atol(long min) { + return tar_atol256(min); +} + +long tar_atosl() { + long n = tar_atol(-1); + if (tar_atosl_min) { + __errno_location(); + return 0; + } + if (n > 0) + return 0; + return n; +} diff --git a/gcc/tree-vect-loop-manip.cc b/gcc/tree-vect-loop-manip.cc index 1066ea17c5674e03412b3dcd8a62ddf4dd54cf31..3810983a80c8b989be9fd9a9993642069fd39b99 100644 --- a/gcc/tree-vect-loop-manip.cc +++ b/gcc/tree-vect-loop-manip.cc @@ -1716,8 +1716,6 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class loop *loop, edge loop_exit, /* Now link the alternative exits. */ if (multiple_exits_p) { - set_immediate_dominator (CDI_DOMINATORS, new_preheader, - main_loop_exit_block); for (auto gsi_from = gsi_start_phis (loop->header), gsi_to = gsi_start_phis (new_preheader); !gsi_end_p (gsi_from) && !gsi_end_p (gsi_to); @@ -1751,12 +1749,26 @@ slpeel_tree_duplicate_loop_to_edge_cfg (class loop *loop, edge loop_exit, /* Finally after wiring the new epilogue we need to update its main exit to the original function exit we recorded. Other exits are already - correct. */ + correct. Because of versioning, skip vectors and others we must update + the dominators of every node reachable by the new exits. */ if (multiple_exits_p) { update_loop = new_loop; - for (edge e : get_loop_exit_edges (loop)) - doms.safe_push (e->dest); + hash_set <basic_block> visited; + auto_vec <edge> workset; + edge ev; + edge_iterator ei; + workset.safe_splice (get_loop_exit_edges (loop)); + while (!workset.is_empty ()) + { + auto bb = workset.pop ()->dest; + if (visited.add (bb)) + continue; + doms.safe_push (bb); + FOR_EACH_EDGE (ev, ei, bb->succs) + workset.safe_push (ev); + } + visited.empty (); doms.safe_push (exit_dest); /* Likely a fall-through edge, so update if needed. */