Mesa (master): nir/lower_io: Add a bounds-checked 64-bit global address format

[GitLab Mirror]

Module: Mesa
Branch: master
Commit: 39da1deb497af55496308c0867b5ce5a0e9df56e
URL:    http://cgit.freedesktop.org/mesa/mesa/commit/?id=39da1deb497af55496308c0867b5ce5a0e9df56e

Author: Jason Ekstrand <jason.ekstrand at intel.com>
Date:   Wed Jan  9 14:56:02 2019 -0600

nir/lower_io: Add a bounds-checked 64-bit global address format

Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin at intel.com>

---

 src/compiler/nir/nir.h          | 10 +++++
 src/compiler/nir/nir_lower_io.c | 89 ++++++++++++++++++++++++++++++++++++++---
 2 files changed, 93 insertions(+), 6 deletions(-)

diff --git a/src/compiler/nir/nir.h b/src/compiler/nir/nir.h
index 92b17275b45..1033b545d6a 100644
--- a/src/compiler/nir/nir.h
+++ b/src/compiler/nir/nir.h
@@ -3038,6 +3038,16 @@ typedef enum {
    nir_address_format_64bit_global,
 
    /**
+    * An address format which is a bounds-checked 64-bit global GPU address.
+    *
+    * The address is comprised as a 32-bit vec4 where .xy are a uint64_t base
+    * address stored with the low bits in .x and high bits in .y, .z is a
+    * size, and .w is an offset.  When the final I/O operation is lowered, .w
+    * is checked against .z and the operation is predicated on the result.
+    */
+   nir_address_format_64bit_bounded_global,
+
+   /**
     * An address format which is comprised of a vec2 where the first
     * component is a buffer index and the second is an offset.
     */
diff --git a/src/compiler/nir/nir_lower_io.c b/src/compiler/nir/nir_lower_io.c
index 749ac91d47e..d39567ff1d1 100644
--- a/src/compiler/nir/nir_lower_io.c
+++ b/src/compiler/nir/nir_lower_io.c
@@ -599,6 +599,13 @@ build_addr_iadd(nir_builder *b, nir_ssa_def *addr,
       assert(addr->num_components == 1);
       return nir_iadd(b, addr, offset);
 
+   case nir_address_format_64bit_bounded_global:
+      assert(addr->num_components == 4);
+      return nir_vec4(b, nir_channel(b, addr, 0),
+                         nir_channel(b, addr, 1),
+                         nir_channel(b, addr, 2),
+                         nir_iadd(b, nir_channel(b, addr, 3), offset));
+
    case nir_address_format_32bit_index_offset:
       assert(addr->num_components == 2);
       return nir_vec2(b, nir_channel(b, addr, 0),
@@ -638,7 +645,8 @@ static bool
 addr_format_is_global(nir_address_format addr_format)
 {
    return addr_format == nir_address_format_32bit_global ||
-          addr_format == nir_address_format_64bit_global;
+          addr_format == nir_address_format_64bit_global ||
+          addr_format == nir_address_format_64bit_bounded_global;
 }
 
 static nir_ssa_def *
@@ -651,6 +659,11 @@ addr_to_global(nir_builder *b, nir_ssa_def *addr,
       assert(addr->num_components == 1);
       return addr;
 
+   case nir_address_format_64bit_bounded_global:
+      assert(addr->num_components == 4);
+      return nir_iadd(b, nir_pack_64_2x32(b, nir_channels(b, addr, 0x3)),
+                         nir_u2u64(b, nir_channel(b, addr, 3)));
+
    case nir_address_format_32bit_index_offset:
       unreachable("Cannot get a 64-bit address with this address format");
    }
@@ -658,6 +671,22 @@ addr_to_global(nir_builder *b, nir_ssa_def *addr,
    unreachable("Invalid address format");
 }
 
+static bool
+addr_format_needs_bounds_check(nir_address_format addr_format)
+{
+   return addr_format == nir_address_format_64bit_bounded_global;
+}
+
+static nir_ssa_def *
+addr_is_in_bounds(nir_builder *b, nir_ssa_def *addr,
+                  nir_address_format addr_format, unsigned size)
+{
+   assert(addr_format == nir_address_format_64bit_bounded_global);
+   assert(addr->num_components == 4);
+   return nir_ige(b, nir_channel(b, addr, 2),
+                     nir_iadd_imm(b, nir_channel(b, addr, 3), size));
+}
+
 static nir_ssa_def *
 build_explicit_io_load(nir_builder *b, nir_intrinsic_instr *intrin,
                        nir_ssa_def *addr, nir_address_format addr_format,
@@ -709,9 +738,32 @@ build_explicit_io_load(nir_builder *b, nir_intrinsic_instr *intrin,
    load->num_components = num_components;
    nir_ssa_dest_init(&load->instr, &load->dest, num_components,
                      intrin->dest.ssa.bit_size, intrin->dest.ssa.name);
-   nir_builder_instr_insert(b, &load->instr);
 
-   return &load->dest.ssa;
+   assert(load->dest.ssa.bit_size % 8 == 0);
+
+   if (addr_format_needs_bounds_check(addr_format)) {
+      /* The Vulkan spec for robustBufferAccess gives us quite a few options
+       * as to what we can do with an OOB read.  Unfortunately, returning
+       * undefined values isn't one of them so we return an actual zero.
+       */
+      nir_const_value zero_val;
+      memset(&zero_val, 0, sizeof(zero_val));
+      nir_ssa_def *zero = nir_build_imm(b, load->num_components,
+                                        load->dest.ssa.bit_size, zero_val);
+
+      const unsigned load_size =
+         (load->dest.ssa.bit_size / 8) * load->num_components;
+      nir_push_if(b, addr_is_in_bounds(b, addr, addr_format, load_size));
+
+      nir_builder_instr_insert(b, &load->instr);
+
+      nir_pop_if(b, NULL);
+
+      return nir_if_phi(b, &load->dest.ssa, zero);
+   } else {
+      nir_builder_instr_insert(b, &load->instr);
+      return &load->dest.ssa;
+   }
 }
 
 static void
@@ -759,7 +811,19 @@ build_explicit_io_store(nir_builder *b, nir_intrinsic_instr *intrin,
    assert(value->num_components == 1 ||
           value->num_components == intrin->num_components);
    store->num_components = value->num_components;
-   nir_builder_instr_insert(b, &store->instr);
+
+   assert(value->bit_size % 8 == 0);
+
+   if (addr_format_needs_bounds_check(addr_format)) {
+      const unsigned store_size = (value->bit_size / 8) * store->num_components;
+      nir_push_if(b, addr_is_in_bounds(b, addr, addr_format, store_size));
+
+      nir_builder_instr_insert(b, &store->instr);
+
+      nir_pop_if(b, NULL);
+   } else {
+      nir_builder_instr_insert(b, &store->instr);
+   }
 }
 
 static nir_ssa_def *
@@ -802,9 +866,22 @@ build_explicit_io_atomic(nir_builder *b, nir_intrinsic_instr *intrin,
    assert(intrin->dest.ssa.num_components == 1);
    nir_ssa_dest_init(&atomic->instr, &atomic->dest,
                      1, intrin->dest.ssa.bit_size, intrin->dest.ssa.name);
-   nir_builder_instr_insert(b, &atomic->instr);
 
-   return &atomic->dest.ssa;
+   assert(atomic->dest.ssa.bit_size % 8 == 0);
+
+   if (addr_format_needs_bounds_check(addr_format)) {
+      const unsigned atomic_size = atomic->dest.ssa.bit_size / 8;
+      nir_push_if(b, addr_is_in_bounds(b, addr, addr_format, atomic_size));
+
+      nir_builder_instr_insert(b, &atomic->instr);
+
+      nir_pop_if(b, NULL);
+      return nir_if_phi(b, &atomic->dest.ssa,
+                           nir_ssa_undef(b, 1, atomic->dest.ssa.bit_size));
+   } else {
+      nir_builder_instr_insert(b, &atomic->instr);
+      return &atomic->dest.ssa;
+   }
 }
 
 static void