Add release notes for the 10.1.5 release.

Update version to 10.1.5
In preparation for the 10.1.5 release, of course.
2014-06-06 19:10:33 -07:00 · 2014-06-06 17:21:44 -07:00 · 2014-06-06 17:14:31 -07:00 · 2014-06-02 13:03:06 -07:00 · 2014-06-02 11:37:13 -07:00 · 2014-06-02 11:35:57 -07:00
15 changed files with 289 additions and 32 deletions
--- a/2
+++ b/2
@@ -1 +1 @@
-10.1.4
+10.1.5
--- a/bin/.cherry-ignore
+++ b/bin/.cherry-ignore
@@ -13,3 +13,13 @@ ac35ded4733883037316d556af596524e5e02535

 # And this one was simply a bug fix for the previously-retracted commit
 2bab95973d8ad3a84f62670143d6f26c230d9582
+
+# Here we have a commit, and its subsequent "revert" both proposed within a
+# single window of the stable release. So we can achieve the same final effect
+# by ignoring both of the commits.
+e3cc0d90e14e62a0a787b6c07a6df0f5c84039be
+0d5ec2c615784929be095951f9269773a790a2dd
+
+# The function being modified here (_eglCreateWindowSurfaceCommon) does not
+# exist in the 10.1 branch.
+91ff0d4c6510dc38f279c586ced17fba917873e7
--- a/docs/relnotes/10.1.4.html
+++ b/docs/relnotes/10.1.4.html
@@ -32,6 +32,9 @@ because compatibility contexts are not supported.

 <h2>MD5 checksums</h2>
 <pre>
+e934365d77f384bfaec844999440bef8  MesaLib-10.1.4.tar.gz
+6fddee101f49b7409cd29994c34ddee7  MesaLib-10.1.4.tar.bz2
+ba5f48e7d5e373922c804c2651fec6c1  MesaLib-10.1.4.zip
 </pre>


--- a/docs/relnotes/10.1.5.html
+++ b/docs/relnotes/10.1.5.html
@@ -0,0 +1,102 @@
+<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
+<html lang="en">
+<head>
+  <meta http-equiv="content-type" content="text/html; charset=utf-8">
+  <title>Mesa Release Notes</title>
+  <link rel="stylesheet" type="text/css" href="../mesa.css">
+</head>
+<body>
+
+<div class="header">
+  <h1>The Mesa 3D Graphics Library</h1>
+</div>
+
+<iframe src="../contents.html"></iframe>
+<div class="content">
+
+<h1>Mesa 10.1.5 Release Notes / (June 6, 2014)</h1>
+
+<p>
+Mesa 10.1.5 is a bug fix release which fixes bugs found since the 10.1.4 release.
+</p>
+
+<p>
+Mesa 10.1.5 implements the OpenGL 3.3 API, but the version reported by
+glGetString(GL_VERSION) or glGetIntegerv(GL_MAJOR_VERSION) /
+glGetIntegerv(GL_MINOR_VERSION) depends on the particular driver being used.
+Some drivers don't support all the features required in OpenGL 3.3.  OpenGL
+3.3 is <strong>only</strong> available if requested at context creation
+because compatibility contexts are not supported.
+</p>
+
+<h2>SHA256 checksums</h2>
+<pre>
+</pre>
+
+<h2>New features</h2>
+<p>None</p>
+
+<h2>Bug fixes</h2>
+
+<p>This list is likely incomplete.</p>
+
+<ul>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=79115">Bug 79115</a> - </li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=79421">Bug 79421</a> - </li>
+
+</ul>
+
+<h2>Changes</h2>
+
+<p>Brian Paul (1):</p>
+<ul>
+  <li>glsl: fix use-after free bug/crash in ast_declarator_list::hir()</li>
+</ul>
+
+<p>Carl Worth (5):</p>
+<ul>
+  <li>docs: Add md5sums for 10.1.4 release</li>
+  <li>Merge remote-tracking branch 'origin/10.1' into 10.1</li>
+  <li>cherry-ignore: Ignore two commits.</li>
+  <li>Ignore a patch that is not needed for the 10.1 branch.</li>
+  <li>Update version to 10.1.5</li>
+</ul>
+
+<p>Emil Velikov (1):</p>
+<ul>
+  <li>glx: do not leak dri3Display</li>
+</ul>
+
+<p>Ilia Mirkin (2):</p>
+<ul>
+  <li>nv50/ir: fix s32 x s32 -&gt; high s32 multiply logic</li>
+  <li>nv50/ir: fix constant folding for OP_MUL subop HIGH</li>
+</ul>
+
+<p>James Legg (1):</p>
+<ul>
+  <li>mesa: Fix unbinding GL_DEPTH_STENCIL_ATTACHMENT</li>
+</ul>
+
+<p>Jeremy Huddleston Sequoia (2):</p>
+<ul>
+  <li>glapi: Avoid heap corruption in _glapi_table</li>
+  <li>darwin: Fix test for kCGLPFAOpenGLProfile support at runtime</li>
+</ul>
+
+<p>Pavel Popov (2):</p>
+<ul>
+  <li>i965: Properly return *RESET* status in glGetGraphicsResetStatusARB</li>
+  <li>i965: Fix Line Stipple enable bit in 3DSTATE_SF for Haswell.</li>
+</ul>
+
+<p>Roland Scheidegger (1):</p>
+<ul>
+  <li>llvmpipe: fix crash when not all attachments are populated in a fb</li>
+</ul>
+
+</div>
+</body>
+</html>
--- a/src/gallium/drivers/llvmpipe/lp_rast.c
+++ b/src/gallium/drivers/llvmpipe/lp_rast.c
@@ -376,9 +376,15 @@ lp_rast_shade_tile(struct lp_rasterizer_task *task,

         /* color buffer */
         for (i = 0; i < scene->fb.nr_cbufs; i++){
-            stride[i] = scene->cbufs[i].stride;
-            color[i] = lp_rast_get_unswizzled_color_block_pointer(task, i, tile_x + x,
-                                                                  tile_y + y, inputs->layer);
+            if (scene->fb.cbufs[i]) {
+               stride[i] = scene->cbufs[i].stride;
+               color[i] = lp_rast_get_unswizzled_color_block_pointer(task, i, tile_x + x,
+                                                                     tile_y + y, inputs->layer);
+            }
+            else {
+               stride[i] = 0;
+               color[i] = NULL;
+            }
         }

         /* depth buffer */
--- a/src/gallium/drivers/nouveau/codegen/nv50_ir_lowering_nv50.cpp
+++ b/src/gallium/drivers/nouveau/codegen/nv50_ir_lowering_nv50.cpp
@@ -37,18 +37,25 @@ namespace nv50_ir {
 //    ah*bl 00
 //
 // fffe0001 + fffe0001
+//
+// Note that this sort of splitting doesn't work for signed values, so we
+// compute the sign on those manually and then perform an unsigned multiply.
 static bool
 expandIntegerMUL(BuildUtil *bld, Instruction *mul)
 {
   const bool highResult = mul->subOp == NV50_IR_SUBOP_MUL_HIGH;

-   DataType fTy = mul->sType; // full type
-   DataType hTy;
+   DataType fTy; // full type
+   switch (mul->sType) {
+   case TYPE_S32: fTy = TYPE_U32; break;
+   case TYPE_S64: fTy = TYPE_U64; break;
+   default: fTy = mul->sType; break;
+   }
+
+   DataType hTy; // half type
   switch (fTy) {
-   case TYPE_S32: hTy = TYPE_S16; break;
   case TYPE_U32: hTy = TYPE_U16; break;
   case TYPE_U64: hTy = TYPE_U32; break;
-   case TYPE_S64: hTy = TYPE_S32; break;
   default:
      return false;
   }
@@ -59,15 +66,25 @@ expandIntegerMUL(BuildUtil *bld, Instruction *mul)

   bld->setPosition(mul, true);

+   Value *s[2];
   Value *a[2], *b[2];
-   Value *c[2];
   Value *t[4];
   for (int j = 0; j < 4; ++j)
      t[j] = bld->getSSA(fullSize);

+   s[0] = mul->getSrc(0);
+   s[1] = mul->getSrc(1);
+
+   if (isSignedType(mul->sType)) {
+      s[0] = bld->getSSA(fullSize);
+      s[1] = bld->getSSA(fullSize);
+      bld->mkOp1(OP_ABS, mul->sType, s[0], mul->getSrc(0));
+      bld->mkOp1(OP_ABS, mul->sType, s[1], mul->getSrc(1));
+   }
+
   // split sources into halves
-   i[0] = bld->mkSplit(a, halfSize, mul->getSrc(0));
-   i[1] = bld->mkSplit(b, halfSize, mul->getSrc(1));
+   i[0] = bld->mkSplit(a, halfSize, s[0]);
+   i[1] = bld->mkSplit(b, halfSize, s[1]);

   i[2] = bld->mkOp2(OP_MUL, fTy, t[0], a[0], b[1]);
   i[3] = bld->mkOp3(OP_MAD, fTy, t[1], a[1], b[0], t[0]);
@@ -75,24 +92,76 @@ expandIntegerMUL(BuildUtil *bld, Instruction *mul)
   i[4] = bld->mkOp3(OP_MAD, fTy, t[3], a[0], b[0], t[2]);

   if (highResult) {
-      Value *r[4];
+      Value *c[2];
+      Value *r[5];
      Value *imm = bld->loadImm(NULL, 1 << (halfSize * 8));
      c[0] = bld->getSSA(1, FILE_FLAGS);
      c[1] = bld->getSSA(1, FILE_FLAGS);
-      for (int j = 0; j < 4; ++j)
+      for (int j = 0; j < 5; ++j)
         r[j] = bld->getSSA(fullSize);

      i[8] = bld->mkOp2(OP_SHR, fTy, r[0], t[1], bld->mkImm(halfSize * 8));
      i[6] = bld->mkOp2(OP_ADD, fTy, r[1], r[0], imm);
      bld->mkMov(r[3], r[0])->setPredicate(CC_NC, c[0]);
      bld->mkOp2(OP_UNION, TYPE_U32, r[2], r[1], r[3]);
-      i[5] = bld->mkOp3(OP_MAD, fTy, mul->getDef(0), a[1], b[1], r[2]);
+      i[5] = bld->mkOp3(OP_MAD, fTy, r[4], a[1], b[1], r[2]);

      // set carry defs / sources
      i[3]->setFlagsDef(1, c[0]);
-      i[4]->setFlagsDef(0, c[1]); // actual result not required, just the carry
+      // actual result required in negative case, but ignored for
+      // unsigned. for some reason the compiler ends up dropping the whole
+      // instruction if the destination is unused but the flags are.
+      if (isSignedType(mul->sType))
+         i[4]->setFlagsDef(1, c[1]);
+      else
+         i[4]->setFlagsDef(0, c[1]);
      i[6]->setPredicate(CC_C, c[0]);
      i[5]->setFlagsSrc(3, c[1]);
+
+      if (isSignedType(mul->sType)) {
+         Value *cc[2];
+         Value *rr[7];
+         Value *one = bld->getSSA(fullSize);
+         bld->loadImm(one, 1);
+         for (int j = 0; j < 7; j++)
+            rr[j] = bld->getSSA(fullSize);
+
+         // NOTE: this logic uses predicates because splitting basic blocks is
+         // ~impossible during the SSA phase. The RA relies on a correlation
+         // between edge order and phi node sources.
+
+         // Set the sign of the result based on the inputs
+         bld->mkOp2(OP_XOR, fTy, NULL, mul->getSrc(0), mul->getSrc(1))
+            ->setFlagsDef(0, (cc[0] = bld->getSSA(1, FILE_FLAGS)));
+
+         // 1s complement of 64-bit value
+         bld->mkOp1(OP_NOT, fTy, rr[0], r[4])
+            ->setPredicate(CC_S, cc[0]);
+         bld->mkOp1(OP_NOT, fTy, rr[1], t[3])
+            ->setPredicate(CC_S, cc[0]);
+
+         // add to low 32-bits, keep track of the carry
+         Instruction *n = bld->mkOp2(OP_ADD, fTy, NULL, rr[1], one);
+         n->setPredicate(CC_S, cc[0]);
+         n->setFlagsDef(0, (cc[1] = bld->getSSA(1, FILE_FLAGS)));
+
+         // If there was a carry, add 1 to the upper 32 bits
+         // XXX: These get executed even if they shouldn't be
+         bld->mkOp2(OP_ADD, fTy, rr[2], rr[0], one)
+            ->setPredicate(CC_C, cc[1]);
+         bld->mkMov(rr[3], rr[0])
+            ->setPredicate(CC_NC, cc[1]);
+         bld->mkOp2(OP_UNION, fTy, rr[4], rr[2], rr[3]);
+
+         // Merge the results from the negative and non-negative paths
+         bld->mkMov(rr[5], rr[4])
+            ->setPredicate(CC_S, cc[0]);
+         bld->mkMov(rr[6], r[4])
+            ->setPredicate(CC_NS, cc[0]);
+         bld->mkOp2(OP_UNION, mul->sType, mul->getDef(0), rr[5], rr[6]);
+      } else {
+         bld->mkMov(mul->getDef(0), r[4]);
+      }
   } else {
      bld->mkMov(mul->getDef(0), t[3]);
   }
--- a/src/gallium/drivers/nouveau/codegen/nv50_ir_peephole.cpp
+++ b/src/gallium/drivers/nouveau/codegen/nv50_ir_peephole.cpp
@@ -418,7 +418,17 @@ ConstantFolding::expr(Instruction *i,
      case TYPE_F32: res.data.f32 = a->data.f32 * b->data.f32; break;
      case TYPE_F64: res.data.f64 = a->data.f64 * b->data.f64; break;
      case TYPE_S32:
-      case TYPE_U32: res.data.u32 = a->data.u32 * b->data.u32; break;
+         if (i->subOp == NV50_IR_SUBOP_MUL_HIGH) {
+            res.data.s32 = ((int64_t)a->data.s32 * b->data.s32) >> 32;
+            break;
+         }
+         /* fallthrough */
+      case TYPE_U32:
+         if (i->subOp == NV50_IR_SUBOP_MUL_HIGH) {
+            res.data.u32 = ((uint64_t)a->data.u32 * b->data.u32) >> 32;
+            break;
+         }
+         res.data.u32 = a->data.u32 * b->data.u32; break;
      default:
         return;
      }
@@ -626,12 +636,41 @@ ConstantFolding::opnd(Instruction *i, ImmediateValue &imm0, int s)
 {
   const int t = !s;
   const operation op = i->op;
+   Instruction *newi = i;

   switch (i->op) {
   case OP_MUL:
      if (i->dType == TYPE_F32)
         tryCollapseChainedMULs(i, s, imm0);

+      if (i->subOp == NV50_IR_SUBOP_MUL_HIGH) {
+         assert(!isFloatType(i->sType));
+         if (imm0.isInteger(1) && i->dType == TYPE_S32) {
+            bld.setPosition(i, false);
+            // Need to set to the sign value, which is a compare.
+            newi = bld.mkCmp(OP_SET, CC_LT, TYPE_S32, i->getDef(0),
+                             TYPE_S32, i->getSrc(t), bld.mkImm(0));
+            delete_Instruction(prog, i);
+         } else if (imm0.isInteger(0) || imm0.isInteger(1)) {
+            // The high bits can't be set in this case (either mul by 0 or
+            // unsigned by 1)
+            i->op = OP_MOV;
+            i->subOp = 0;
+            i->setSrc(0, new_ImmediateValue(prog, 0u));
+            i->src(0).mod = Modifier(0);
+            i->setSrc(1, NULL);
+         } else if (!imm0.isNegative() && imm0.isPow2()) {
+            // Translate into a shift
+            imm0.applyLog2();
+            i->op = OP_SHR;
+            i->subOp = 0;
+            imm0.reg.data.u32 = 32 - imm0.reg.data.u32;
+            i->setSrc(0, i->getSrc(t));
+            i->src(0).mod = i->src(t).mod;
+            i->setSrc(1, new_ImmediateValue(prog, imm0.reg.data.u32));
+            i->src(1).mod = 0;
+         }
+      } else
      if (imm0.isInteger(0)) {
         i->op = OP_MOV;
         i->setSrc(0, new_ImmediateValue(prog, 0u));
@@ -722,7 +761,7 @@ ConstantFolding::opnd(Instruction *i, ImmediateValue &imm0, int s)
         else
            tA = tB;
         tB = s ? bld.getSSA() : i->getDef(0);
-         bld.mkOp2(OP_ADD, TYPE_U32, tB, mul->getDef(0), tA);
+         newi = bld.mkOp2(OP_ADD, TYPE_U32, tB, mul->getDef(0), tA);
         if (s)
            bld.mkOp2(OP_SHR, TYPE_U32, i->getDef(0), tB, bld.mkImm(s));

@@ -754,7 +793,7 @@ ConstantFolding::opnd(Instruction *i, ImmediateValue &imm0, int s)
         tA = bld.getSSA();
         bld.mkCmp(OP_SET, CC_LT, TYPE_S32, tA, TYPE_S32, i->getSrc(0), bld.mkImm(0));
         tD = (d < 0) ? bld.getSSA() : i->getDef(0)->asLValue();
-         bld.mkOp2(OP_SUB, TYPE_U32, tD, tB, tA);
+         newi = bld.mkOp2(OP_SUB, TYPE_U32, tD, tB, tA);
         if (d < 0)
            bld.mkOp1(OP_NEG, TYPE_S32, i->getDef(0), tB);

@@ -832,7 +871,7 @@ ConstantFolding::opnd(Instruction *i, ImmediateValue &imm0, int s)
   default:
      return;
   }
-   if (i->op != op)
+   if (newi->op != op)
      foldCount++;
 }

--- a/src/gallium/drivers/nouveau/codegen/nv50_ir_target_nv50.cpp
+++ b/src/gallium/drivers/nouveau/codegen/nv50_ir_target_nv50.cpp
@@ -329,6 +329,8 @@ TargetNV50::insnCanLoad(const Instruction *i, int s,
         return false;
      if (sf == FILE_IMMEDIATE)
         return false;
+      if (i->subOp == NV50_IR_SUBOP_MUL_HIGH && sf == FILE_MEMORY_CONST)
+         return false;
      ldSize = 2;
   } else {
      ldSize = typeSizeof(ld->dType);
--- a/src/glsl/ast_to_hir.cpp
+++ b/src/glsl/ast_to_hir.cpp
@@ -3494,11 +3494,15 @@ ast_declarator_list::hir(exec_list *instructions,
       * instruction stream.
       */
      exec_list initializer_instructions;
+
+      /* Examine var name here since var may get deleted in the next call */
+      bool var_is_gl_id = (strncmp(var->name, "gl_", 3) == 0);
+
      ir_variable *earlier =
         get_variable_being_redeclared(var, decl->get_location(), state,
                                       false /* allow_all_redeclarations */);
      if (earlier != NULL) {
-         if (strncmp(var->name, "gl_", 3) == 0 &&
+         if (var_is_gl_id &&
             earlier->data.how_declared == ir_var_declared_in_block) {
            _mesa_glsl_error(&loc, state,
                             "`%s' has already been redeclared using "
--- a/src/glx/apple/apple_visual.c
+++ b/src/glx/apple/apple_visual.c
@@ -73,11 +73,15 @@ apple_visual_create_pfobj(CGLPixelFormatObj * pfobj, const struct glx_config * m
   GLint vsref = 0;
   CGLError error = 0;

-   /* Request an OpenGL 3.2 profile if one is available */
-   if(apple_cgl.version_major > 1 || (apple_cgl.version_major == 1 && apple_cgl.version_minor >= 3)) {
-      attr[numattr++] = kCGLPFAOpenGLProfile;
-      attr[numattr++] = kCGLOGLPVersion_3_2_Core;
-   }
+   /* Request an OpenGL 3.2 profile if one is available and supported */
+   attr[numattr++] = kCGLPFAOpenGLProfile;
+   attr[numattr++] = kCGLOGLPVersion_3_2_Core;
+
+   /* Test for kCGLPFAOpenGLProfile support at runtime and roll it out if not supported */
+   attr[numattr] = 0;
+   error = apple_cgl.choose_pixel_format(attr, pfobj, &vsref);
+   if (error == kCGLBadAttribute)
+      numattr -= 2;

   if (offscreen) {
      apple_glx_diagnostic
--- a/src/glx/glxext.c
+++ b/src/glx/glxext.c
@@ -249,6 +249,10 @@ glx_display_free(struct glx_display *priv)
   if (priv->dri2Display)
      (*priv->dri2Display->destroyDisplay) (priv->dri2Display);
   priv->dri2Display = NULL;
+
+   if (priv->dri3Display)
+      (*priv->dri3Display->destroyDisplay) (priv->dri3Display);
+   priv->dri3Display = NULL;
 #endif

   free((char *) priv);
--- a/src/mapi/glapi/gen/gl_gentable.py
+++ b/src/mapi/glapi/gen/gl_gentable.py
@@ -113,7 +113,7 @@ __glapi_gentable_set_remaining_noop(struct _glapi_table *disp) {

 struct _glapi_table *
 _glapi_create_table_from_handle(void *handle, const char *symbol_prefix) {
-    struct _glapi_table *disp = calloc(1, sizeof(struct _glapi_table));
+    struct _glapi_table *disp = calloc(1, _glapi_get_dispatch_table_size() * sizeof(_glapi_proc));
    char symboln[512];

    if(!disp)
--- a/src/mesa/drivers/dri/i965/brw_defines.h
+++ b/src/mesa/drivers/dri/i965/brw_defines.h
@@ -1676,7 +1676,7 @@ enum brw_message_target {
 /* GEN7/DW1: */
 # define GEN7_SF_DEPTH_BUFFER_SURFACE_FORMAT_SHIFT	12
 /* GEN7/DW2: */
-# define HSW_SF_LINE_STIPPLE_ENABLE			14
+# define HSW_SF_LINE_STIPPLE_ENABLE			(1 << 14)

 # define GEN8_SF_SMOOTH_POINT_ENABLE                    (1 << 13)

--- a/src/mesa/drivers/dri/i965/brw_reset.c
+++ b/src/mesa/drivers/dri/i965/brw_reset.c
@@ -42,6 +42,13 @@ brw_get_graphics_reset_status(struct gl_context *ctx)
    */
   assert(brw->hw_ctx != NULL);

+   /* A reset status other than NO_ERROR was returned last time. I915 returns
+    * nonzero active/pending only if reset has been encountered and completed.
+    * Return NO_ERROR from now on.
+    */
+   if (brw->reset_count != 0)
+      return GL_NO_ERROR;
+
   err = drm_intel_get_reset_stats(brw->hw_ctx, &reset_count, &active,
                                   &pending);
   if (err)
@@ -50,18 +57,19 @@ brw_get_graphics_reset_status(struct gl_context *ctx)
   /* A reset was observed while a batch from this context was executing.
    * Assume that this context was at fault.
    */
-   if (active != 0)
+   if (active != 0) {
+      brw->reset_count = reset_count;
      return GL_GUILTY_CONTEXT_RESET_ARB;
+   }

   /* A reset was observed while a batch from this context was in progress,
    * but the batch was not executing.  In this case, assume that the context
    * was not at fault.
    */
-   if (pending != 0)
+   if (pending != 0) {
+      brw->reset_count = reset_count;
      return GL_INNOCENT_CONTEXT_RESET_ARB;
-
-   /* FINISHME: Should we report anything if reset_count > brw->reset_count?
-    */
+   }

   return GL_NO_ERROR;
 }
--- a/src/mesa/main/fbobject.c
+++ b/src/mesa/main/fbobject.c
@@ -500,6 +500,12 @@ _mesa_framebuffer_renderbuffer(struct gl_context *ctx,
   }
   else {
      remove_attachment(ctx, att);
+      if (attachment == GL_DEPTH_STENCIL_ATTACHMENT) {
+         /* detach stencil (depth was detached above) */
+         att = get_attachment(ctx, fb, GL_STENCIL_ATTACHMENT_EXT);
+         assert(att);
+         remove_attachment(ctx, att);
+      }
   }

   invalidate_framebuffer(fb);
Author	SHA1	Message	Date
Carl Worth	feb4c7284c	Add release notes for the 10.1.5 release.	2014-06-06 19:10:33 -07:00
Carl Worth	b614628a3c	Update version to 10.1.5 In preparation for the 10.1.5 release, of course.	2014-06-06 17:21:44 -07:00
Carl Worth	1f08d1bf46	Ignore a patch that is not needed for the 10.1 branch. The function being modified does not exist in 10.1.	2014-06-06 17:14:31 -07:00
Carl Worth	a73894a7ed	cherry-ignore: Ignore two commits. The second of these two is simply a "git revert" of the first. So skipping both of them gives us the same final result in a simpler way.	2014-06-02 13:03:06 -07:00
Roland Scheidegger	cf08c24750	llvmpipe: fix crash when not all attachments are populated in a fb Framebuffers can have NULL attachments since a while. llvmpipe handled that properly for lp_rast_shade_quads_mask but it seems the change didn't make it to lp_rast_shade_tile. This fixes piglit fbo-drawbuffers-none test (though I need to increase the FB_SIZE from 32 to 256 so the tris cover some tiles fully). https://bugs.freedesktop.org/show_bug.cgi?id=79421 Cc: "10.1 10.2" <mesa-stable@lists.freedesktop.org> Reviewed-by: Jose Fonseca <jfonseca@vmware.com> (cherry picked from commit `576868140b`)	2014-06-02 11:37:13 -07:00
Pavel Popov	4d676c5ed2	i965: Fix Line Stipple enable bit in 3DSTATE_SF for Haswell. Cc: "10.1 10.2" <mesa-stable@lists.freedesktop.org> Reviewed-by: Matt Turner <mattst88@gmail.com> Signed-off-by: Pavel Popov <pavel.e.popov@intel.com> (cherry picked from commit `d292d40207`)	2014-06-02 11:35:57 -07:00
Brian Paul	4942eae869	glsl: fix use-after free bug/crash in ast_declarator_list::hir() The call to get_variable_being_redeclared() may delete 'var' so we can't reference var->name afterward. We fix that by examining the var's name before making that call. Fixes valgrind warnings and possible crash when running the piglit tests/spec/glsl-1.30/execution/clipping/vs-clip-distance-in-param.shader_test test (and probably others). Cc: "10.1 10.2" <mesa-stable@lists.freedesktop.org> Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> (cherry picked from commit `f9cecca7a6`)	2014-06-02 11:31:55 -07:00
Emil Velikov	1776a562b4	glx: do not leak dri3Display v2: Do not wrap the code in ifdef HAVE_DRI3 (suggested by Keith) Cc: "10.1 10.2" <mesa-stable@lists.freedesktop.org> Cc: Keith Packard <keithp@keithp.com> Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> (cherry picked from commit `eb2241f8a9`)	2014-06-02 11:30:15 -07:00
Pavel Popov	d2f5638ade	i965: Properly return RESET status in glGetGraphicsResetStatusARB The glGetGraphicsResetStatusARB from ARB_robustness extension always returns GUILTY_CONTEXT_RESET_ARB and never returns NO_ERROR for guilty context with LOSE_CONTEXT_ON_RESET_ARB strategy. This is because Mesa returns GUILTY_CONTEXT_RESET_ARB if batch_active !=0 whereas kernel driver never reset batch_active and this variable always > 0 for guilty context. The same behaviour also can be observed for batch_pending and INNOCENT_CONTEXT_RESET_ARB. But ARB_robustness spec says: If a reset status other than NO_ERROR is returned and subsequent calls return NO_ERROR, the context reset was encountered and completed. If a reset status is repeatedly returned, the context may be in the process of resetting. 8. How should the application react to a reset context event? RESOLVED: For this extension, the application is expected to query the reset status until NO_ERROR is returned. If a reset is encountered, at least one RESET status will be returned. Once NO_ERROR is encountered, the application can safely destroy the old context and create a new one. The main problem is the context may be in the process of resetting and in this case a reset status should be repeatedly returned. But looks like the kernel driver returns nonzero active/pending only if the context reset has already been encountered and completed. For this reason the RESET status cannot be repeatedly returned and should be returned only once. The reset_count and brw->reset_count variables can be used to control that glGetGraphicsResetStatusARB returns RESET status only once for each context. Note the i915 triggers reset_count twice which allows to return correct reset count immediately after active/pending have been incremented. v2 (idr): Trivial reformatting of comments. Signed-off-by: Pavel Popov <pavel.e.popov@intel.com> Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> Cc: "10.1 10.2" <mesa-stable@lists.freedesktop.org> (cherry picked from commit `8dc4a98c44`)	2014-06-02 11:28:46 -07:00
James Legg	ee0207a212	mesa: Fix unbinding GL_DEPTH_STENCIL_ATTACHMENT glFramebufferRender(..., GL_DEPTH_STENCIL_ATTACHMENT, ..., 0) only detached the depth buffer and not the stencil buffer. Bugzilla: http://bugs.freedesktop.org/show_bug.cgi?id=79115 Reviewed-by: Brian Paul <brianp@vmware.com> Cc: "10.1 10.2" <mesa-stable@lists.freedesktop.org> (cherry picked from commit `846c715abb`)	2014-06-02 11:25:24 -07:00
Ilia Mirkin	36e0e9c5e7	nv50/ir: fix constant folding for OP_MUL subop HIGH These instructions can come in either through IMUL_HI/UMUL_HI TGSI opcodes, or from OP_DIV constant folding. Also make sure that the constant foldings which delete the original instruction still get counted as having done something. Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> Cc: "10.1 10.2" <mesa-stable@lists.freedesktop.org> Reviewed-by: Ben Skeggs <bskeggs@redhat.com> (cherry picked from commit `d2a3de19c6`)	2014-05-22 11:33:40 -07:00
Ilia Mirkin	5d8e60dcc7	nv50/ir: fix s32 x s32 -> high s32 multiply logic Retrieving the high 32 bits of a signed multiply is rather annoying. It appears that the simplest way to do this is to compute the absolute value of the arguments, and perform a u32 x u32 -> u64 operation. If the arguments' signs differ, then negate the result. Since there is no u64 support in the cvt instruction, we have the perform the 2's complement negation "by hand". This logic can come into use by the IMUL_HI instruction (very unlikely to be seen), as well as from constant folding of division by a constant. Fixes dolphin's divisions by 255. Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> Cc: "10.1 10.2" <mesa-stable@lists.freedesktop.org> Reviewed-by: Ben Skeggs <bskeggs@redhat.com> (cherry picked from commit `d3a5cf052c`)	2014-05-22 11:32:51 -07:00
Carl Worth	a23e73e00d	Merge remote-tracking branch 'origin/10.1' into 10.1	2014-05-20 15:28:57 -07:00
Carl Worth	a02f6639f7	docs: Add md5sums for 10.1.4 release After making the tar files.	2014-05-20 15:26:00 -07:00
Jeremy Huddleston Sequoia	ec83a39e2b	darwin: Fix test for kCGLPFAOpenGLProfile support at runtime Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> (cherry picked from commit `7a109268ab`)	2014-05-20 10:55:27 -07:00
Jeremy Huddleston Sequoia	ea5839c8fe	glapi: Avoid heap corruption in _glapi_table Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> Reviewed-by: Chia-I Wu <olv@lunarg.com> (cherry picked from commit `ff5456d1ac`)	2014-05-20 01:39:32 -07:00
@@ -1 +1 @@
 .1.4
 .1.5