Merge tag 'pull-tcg-20230731' of https://gitlab.com/rth7680/qemu into staging

util/interval-tree: Access left/right/parent atomically
accel/tcg: Clear gen_tb on buffer overflow
bsd-user: Specify host page alignment if none specified
bsd-user: Allocate guest virtual address space
target/ppc: Disable goto_tb with architectural singlestep
target/s390x: Move trans_exc_code update to do_program_interrupt

# -----BEGIN PGP SIGNATURE-----
#
# iQFRBAABCgA7FiEEekgeeIaLTbaoWgXAZN846K9+IV8FAmTIIQUdHHJpY2hhcmQu
# aGVuZGVyc29uQGxpbmFyby5vcmcACgkQZN846K9+IV87JAf/ZgJTq26oniJ4TLkS
# 2UVBEcxGnnA2L1n4zcXG1o0onT5dAqm/6YjSlVD7C+Ol8pzQMomJKcWLL/jrCEUp
# rQXPV9ibD5bCtO47MY3ZS3aW3pqOhXOeKUFer1+YHWRRyi9Y6kEx0d2No3MSGo18
# S5A6zPwqduQvZPBPVualmtdIrpTasxhUdNfbqBW31pxYpCNg1wqIiwKoLcD5NJeX
# epVhaUi/7TwqljrK7SGXmmfDWiTHIXDtvPrJQcSYGgqpVNFzRuq6jTXRJObeWen0
# DhOHqC0Z6OkZ2gU+eso/VRbcbawQNQohUHQzZ7c0643TxncPDKG82/MDRe2MTJnq
# /z+jpw==
# =Z8UY
# -----END PGP SIGNATURE-----
# gpg: Signature made Mon 31 Jul 2023 02:00:53 PM PDT
# gpg:                using RSA key 7A481E78868B4DB6A85A05C064DF38E8AF7E215F
# gpg:                issuer "richard.henderson@linaro.org"
# gpg: Good signature from "Richard Henderson <richard.henderson@linaro.org>" [ultimate]

* tag 'pull-tcg-20230731' of https://gitlab.com/rth7680/qemu:
  target/s390x: Move trans_exc_code update to do_program_interrupt
  linux-user/armeb: Fix __kernel_cmpxchg() for armeb
  target/ppc: Disable goto_tb with architectural singlestep
  bsd-user: Specify host page alignment if none specified
  bsd-user: Allocate guest virtual address space
  accel/tcg: Clear tcg_ctx->gen_tb on buffer overflow
  util/interval-tree: Use qatomic_read/set for rb_parent_color
  util/interval-tree: Introduce pc_parent
  util/interval-tree: Use qatomic_set_mb in rb_link_node
  util/interval-tree: Use qatomic_read for left/right while searching

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

7 files changed, 132 insertions, 51 deletions

diff --git a/accel/tcg/translate-all.c b/accel/tcg/translate-all.c
index a1782db5dd..b2d4e22c17 100644
--- a/accel/tcg/translate-all.c
+++ b/accel/tcg/translate-all.c
@@ -374,6 +374,7 @@ TranslationBlock *tb_gen_code(CPUState *cpu,
                           "Restarting code generation for "
                           "code_gen_buffer overflow\n");
             tb_unlock_pages(tb);
+            tcg_ctx->gen_tb = NULL;
             goto buffer_overflow;
 
         case -2:
diff --git a/bsd-user/main.c b/bsd-user/main.c
index b597328118..381bb18df8 100644
--- a/bsd-user/main.c
+++ b/bsd-user/main.c
@@ -473,10 +473,6 @@ int main(int argc, char **argv)
     target_environ = envlist_to_environ(envlist, NULL);
     envlist_free(envlist);
 
-    if (reserved_va) {
-        mmap_next_start = reserved_va + 1;
-    }
-
     {
         Error *err = NULL;
         if (seed_optarg != NULL) {
@@ -494,7 +490,49 @@ int main(int argc, char **argv)
      * Now that page sizes are configured we can do
      * proper page alignment for guest_base.
      */
-    guest_base = HOST_PAGE_ALIGN(guest_base);
+    if (have_guest_base) {
+        if (guest_base & ~qemu_host_page_mask) {
+            error_report("Selected guest base not host page aligned");
+            exit(1);
+        }
+    }
+
+    /*
+     * If reserving host virtual address space, do so now.
+     * Combined with '-B', ensure that the chosen range is free.
+     */
+    if (reserved_va) {
+        void *p;
+
+        if (have_guest_base) {
+            p = mmap((void *)guest_base, reserved_va + 1, PROT_NONE,
+                     MAP_ANON | MAP_PRIVATE | MAP_FIXED | MAP_EXCL, -1, 0);
+        } else {
+            p = mmap(NULL, reserved_va + 1, PROT_NONE,
+                     MAP_ANON | MAP_PRIVATE, -1, 0);
+        }
+        if (p == MAP_FAILED) {
+            const char *err = strerror(errno);
+            char *sz = size_to_str(reserved_va + 1);
+
+            if (have_guest_base) {
+                error_report("Cannot allocate %s bytes at -B %p for guest "
+                             "address space: %s", sz, (void *)guest_base, err);
+            } else {
+                error_report("Cannot allocate %s bytes for guest "
+                             "address space: %s", sz, err);
+            }
+            exit(1);
+        }
+        guest_base = (uintptr_t)p;
+        have_guest_base = true;
+
+        /* Ensure that mmap_next_start is within range. */
+        if (reserved_va <= mmap_next_start) {
+            mmap_next_start = (reserved_va / 4 * 3)
+                              & TARGET_PAGE_MASK & qemu_host_page_mask;
+        }
+    }
 
     if (loader_exec(filename, argv + optind, target_environ, regs, info,
                     &bprm) != 0) {
diff --git a/bsd-user/mmap.c b/bsd-user/mmap.c
index 74ed00b9fe..b62a69bd07 100644
--- a/bsd-user/mmap.c
+++ b/bsd-user/mmap.c
@@ -260,7 +260,8 @@ static abi_ulong mmap_find_vma_aligned(abi_ulong start, abi_ulong size,
 
     if (reserved_va) {
         return mmap_find_vma_reserved(start, size,
-            (alignment != 0 ? 1 << alignment : 0));
+            (alignment != 0 ? 1 << alignment :
+             MAX(qemu_host_page_size, TARGET_PAGE_SIZE)));
     }
 
     addr = start;
diff --git a/linux-user/arm/cpu_loop.c b/linux-user/arm/cpu_loop.c
index a992423257..b404117ff3 100644
--- a/linux-user/arm/cpu_loop.c
+++ b/linux-user/arm/cpu_loop.c
@@ -117,8 +117,9 @@ static void arm_kernel_cmpxchg32_helper(CPUARMState *env)
 {
     uint32_t oldval, newval, val, addr, cpsr, *host_addr;
 
-    oldval = env->regs[0];
-    newval = env->regs[1];
+    /* Swap if host != guest endianness, for the host cmpxchg below */
+    oldval = tswap32(env->regs[0]);
+    newval = tswap32(env->regs[1]);
     addr = env->regs[2];
 
     mmap_lock();
@@ -174,6 +175,10 @@ static void arm_kernel_cmpxchg64_helper(CPUARMState *env)
         return;
     }
 
+    /* Swap if host != guest endianness, for the host cmpxchg below */
+    oldval = tswap64(oldval);
+    newval = tswap64(newval);
+
 #ifdef CONFIG_ATOMIC64
     val = qatomic_cmpxchg__nocheck(host_addr, oldval, newval);
     cpsr = (val == oldval) * CPSR_C;
diff --git a/target/ppc/translate.c b/target/ppc/translate.c
index e6a0709066..74796ec7ba 100644
--- a/target/ppc/translate.c
+++ b/target/ppc/translate.c
@@ -4175,6 +4175,9 @@ static void pmu_count_insns(DisasContext *ctx)
 
 static inline bool use_goto_tb(DisasContext *ctx, target_ulong dest)
 {
+    if (unlikely(ctx->singlestep_enabled)) {
+        return false;
+    }
     return translator_use_goto_tb(&ctx->base, dest);
 }
 
diff --git a/target/s390x/tcg/excp_helper.c b/target/s390x/tcg/excp_helper.c
index 3da337f7c7..b7116d0577 100644
--- a/target/s390x/tcg/excp_helper.c
+++ b/target/s390x/tcg/excp_helper.c
@@ -190,11 +190,6 @@ bool s390_cpu_tlb_fill(CPUState *cs, vaddr address, int size,
         return false;
     }
 
-    if (excp != PGM_ADDRESSING) {
-        stq_phys(env_cpu(env)->as,
-                 env->psa + offsetof(LowCore, trans_exc_code), tec);
-    }
-
     /*
      * For data accesses, ILEN will be filled in from the unwind info,
      * within cpu_loop_exit_restore.  For code accesses, retaddr == 0,
@@ -211,20 +206,33 @@ static void do_program_interrupt(CPUS390XState *env)
     uint64_t mask, addr;
     LowCore *lowcore;
     int ilen = env->int_pgm_ilen;
+    bool set_trans_exc_code = false;
+    bool advance = false;
 
     assert((env->int_pgm_code == PGM_SPECIFICATION && ilen == 0) ||
            ilen == 2 || ilen == 4 || ilen == 6);
 
     switch (env->int_pgm_code) {
     case PGM_PER:
-        if (env->per_perc_atmid & PER_CODE_EVENT_NULLIFICATION) {
-            break;
-        }
-        /* FALL THROUGH */
+        advance = !(env->per_perc_atmid & PER_CODE_EVENT_NULLIFICATION);
+        break;
+    case PGM_ASCE_TYPE:
+    case PGM_REG_FIRST_TRANS:
+    case PGM_REG_SEC_TRANS:
+    case PGM_REG_THIRD_TRANS:
+    case PGM_SEGMENT_TRANS:
+    case PGM_PAGE_TRANS:
+        assert(env->int_pgm_code == env->tlb_fill_exc);
+        set_trans_exc_code = true;
+        break;
+    case PGM_PROTECTION:
+        assert(env->int_pgm_code == env->tlb_fill_exc);
+        set_trans_exc_code = true;
+        advance = true;
+        break;
     case PGM_OPERATION:
     case PGM_PRIVILEGED:
     case PGM_EXECUTE:
-    case PGM_PROTECTION:
     case PGM_ADDRESSING:
     case PGM_SPECIFICATION:
     case PGM_DATA:
@@ -243,11 +251,15 @@ static void do_program_interrupt(CPUS390XState *env)
     case PGM_PC_TRANS_SPEC:
     case PGM_ALET_SPEC:
     case PGM_MONITOR:
-        /* advance the PSW if our exception is not nullifying */
-        env->psw.addr += ilen;
+        advance = true;
         break;
     }
 
+    /* advance the PSW if our exception is not nullifying */
+    if (advance) {
+        env->psw.addr += ilen;
+    }
+
     qemu_log_mask(CPU_LOG_INT,
                   "%s: code=0x%x ilen=%d psw: %" PRIx64 " %" PRIx64 "\n",
                   __func__, env->int_pgm_code, ilen, env->psw.mask,
@@ -263,6 +275,10 @@ static void do_program_interrupt(CPUS390XState *env)
         env->per_perc_atmid = 0;
     }
 
+    if (set_trans_exc_code) {
+        lowcore->trans_exc_code = cpu_to_be64(env->tlb_fill_tec);
+    }
+
     lowcore->pgm_ilen = cpu_to_be16(ilen);
     lowcore->pgm_code = cpu_to_be16(env->int_pgm_code);
     lowcore->program_old_psw.mask = cpu_to_be64(s390_cpu_get_psw_mask(env));
diff --git a/util/interval-tree.c b/util/interval-tree.c
index 4c0baf108f..f2866aa7d3 100644
--- a/util/interval-tree.c
+++ b/util/interval-tree.c
@@ -48,12 +48,6 @@
  *
  * It also guarantees that if the lookup returns an element it is the 'correct'
  * one. But not returning an element does _NOT_ mean it's not present.
- *
- * NOTE:
- *
- * Stores to __rb_parent_color are not important for simple lookups so those
- * are left undone as of now. Nor did I check for loops involving parent
- * pointers.
  */
 
 typedef enum RBColor
@@ -68,14 +62,29 @@ typedef struct RBAugmentCallbacks {
     void (*rotate)(RBNode *old, RBNode *new);
 } RBAugmentCallbacks;
 
+static inline uintptr_t rb_pc(const RBNode *n)
+{
+    return qatomic_read(&n->rb_parent_color);
+}
+
+static inline void rb_set_pc(RBNode *n, uintptr_t pc)
+{
+    qatomic_set(&n->rb_parent_color, pc);
+}
+
+static inline RBNode *pc_parent(uintptr_t pc)
+{
+    return (RBNode *)(pc & ~1);
+}
+
 static inline RBNode *rb_parent(const RBNode *n)
 {
-    return (RBNode *)(n->rb_parent_color & ~1);
+    return pc_parent(rb_pc(n));
 }
 
 static inline RBNode *rb_red_parent(const RBNode *n)
 {
-    return (RBNode *)n->rb_parent_color;
+    return (RBNode *)rb_pc(n);
 }
 
 static inline RBColor pc_color(uintptr_t pc)
@@ -95,27 +104,27 @@ static inline bool pc_is_black(uintptr_t pc)
 
 static inline RBColor rb_color(const RBNode *n)
 {
-    return pc_color(n->rb_parent_color);
+    return pc_color(rb_pc(n));
 }
 
 static inline bool rb_is_red(const RBNode *n)
 {
-    return pc_is_red(n->rb_parent_color);
+    return pc_is_red(rb_pc(n));
 }
 
 static inline bool rb_is_black(const RBNode *n)
 {
-    return pc_is_black(n->rb_parent_color);
+    return pc_is_black(rb_pc(n));
 }
 
 static inline void rb_set_black(RBNode *n)
 {
-    n->rb_parent_color |= RB_BLACK;
+    rb_set_pc(n, rb_pc(n) | RB_BLACK);
 }
 
 static inline void rb_set_parent_color(RBNode *n, RBNode *p, RBColor color)
 {
-    n->rb_parent_color = (uintptr_t)p | color;
+    rb_set_pc(n, (uintptr_t)p | color);
 }
 
 static inline void rb_set_parent(RBNode *n, RBNode *p)
@@ -128,7 +137,11 @@ static inline void rb_link_node(RBNode *node, RBNode *parent, RBNode **rb_link)
     node->rb_parent_color = (uintptr_t)parent;
     node->rb_left = node->rb_right = NULL;
 
-    qatomic_set(rb_link, node);
+    /*
+     * Ensure that node is initialized before insertion,
+     * as viewed by a concurrent search.
+     */
+    qatomic_set_mb(rb_link, node);
 }
 
 static RBNode *rb_next(RBNode *node)
@@ -177,9 +190,10 @@ static inline void rb_change_child(RBNode *old, RBNode *new,
 static inline void rb_rotate_set_parents(RBNode *old, RBNode *new,
                                          RBRoot *root, RBColor color)
 {
-    RBNode *parent = rb_parent(old);
+    uintptr_t pc = rb_pc(old);
+    RBNode *parent = pc_parent(pc);
 
-    new->rb_parent_color = old->rb_parent_color;
+    rb_set_pc(new, pc);
     rb_set_parent_color(old, new, color);
     rb_change_child(old, new, parent, root);
 }
@@ -527,11 +541,11 @@ static void rb_erase_augmented(RBNode *node, RBRoot *root,
          * and node must be black due to 4). We adjust colors locally
          * so as to bypass rb_erase_color() later on.
          */
-        pc = node->rb_parent_color;
-        parent = rb_parent(node);
+        pc = rb_pc(node);
+        parent = pc_parent(pc);
         rb_change_child(node, child, parent, root);
         if (child) {
-            child->rb_parent_color = pc;
+            rb_set_pc(child, pc);
             rebalance = NULL;
         } else {
             rebalance = pc_is_black(pc) ? parent : NULL;
@@ -539,9 +553,9 @@ static void rb_erase_augmented(RBNode *node, RBRoot *root,
         tmp = parent;
     } else if (!child) {
         /* Still case 1, but this time the child is node->rb_left */
-        pc = node->rb_parent_color;
-        parent = rb_parent(node);
-        tmp->rb_parent_color = pc;
+        pc = rb_pc(node);
+        parent = pc_parent(pc);
+        rb_set_pc(tmp, pc);
         rb_change_child(node, tmp, parent, root);
         rebalance = NULL;
         tmp = parent;
@@ -595,8 +609,8 @@ static void rb_erase_augmented(RBNode *node, RBRoot *root,
         qatomic_set(&successor->rb_left, tmp);
         rb_set_parent(tmp, successor);
 
-        pc = node->rb_parent_color;
-        tmp = rb_parent(node);
+        pc = rb_pc(node);
+        tmp = pc_parent(pc);
         rb_change_child(node, successor, tmp, root);
 
         if (child2) {
@@ -605,7 +619,7 @@ static void rb_erase_augmented(RBNode *node, RBRoot *root,
         } else {
             rebalance = rb_is_black(successor) ? parent : NULL;
         }
-        successor->rb_parent_color = pc;
+        rb_set_pc(successor, pc);
         tmp = successor;
     }
 
@@ -745,8 +759,9 @@ static IntervalTreeNode *interval_tree_subtree_search(IntervalTreeNode *node,
          * Loop invariant: start <= node->subtree_last
          * (Cond2 is satisfied by one of the subtree nodes)
          */
-        if (node->rb.rb_left) {
-            IntervalTreeNode *left = rb_to_itree(node->rb.rb_left);
+        RBNode *tmp = qatomic_read(&node->rb.rb_left);
+        if (tmp) {
+            IntervalTreeNode *left = rb_to_itree(tmp);
 
             if (start <= left->subtree_last) {
                 /*
@@ -765,8 +780,9 @@ static IntervalTreeNode *interval_tree_subtree_search(IntervalTreeNode *node,
             if (start <= node->last) {     /* Cond2 */
                 return node; /* node is leftmost match */
             }
-            if (node->rb.rb_right) {
-                node = rb_to_itree(node->rb.rb_right);
+            tmp = qatomic_read(&node->rb.rb_right);
+            if (tmp) {
+                node = rb_to_itree(tmp);
                 if (start <= node->subtree_last) {
                     continue;
                 }
@@ -814,8 +830,9 @@ IntervalTreeNode *interval_tree_iter_first(IntervalTreeRoot *root,
 IntervalTreeNode *interval_tree_iter_next(IntervalTreeNode *node,
                                           uint64_t start, uint64_t last)
 {
-    RBNode *rb = node->rb.rb_right, *prev;
+    RBNode *rb, *prev;
 
+    rb = qatomic_read(&node->rb.rb_right);
     while (true) {
         /*
          * Loop invariants:
@@ -840,7 +857,7 @@ IntervalTreeNode *interval_tree_iter_next(IntervalTreeNode *node,
             }
             prev = &node->rb;
             node = rb_to_itree(rb);
-            rb = node->rb.rb_right;
+            rb = qatomic_read(&node->rb.rb_right);
         } while (prev == rb);
 
         /* Check if the node intersects [start;last] */