qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Richard Henderson <richard.henderson@linaro.org>
To: qemu-devel@nongnu.org
Subject: [PATCH 05/45] target/hppa: Allow prior nullification in do_ibranch
Date: Wed, 24 Apr 2024 16:59:43 -0700	[thread overview]
Message-ID: <20240425000023.1002026-6-richard.henderson@linaro.org> (raw)
In-Reply-To: <20240425000023.1002026-1-richard.henderson@linaro.org>

Simplify the function by not attempting a conditional move
on the branch destination -- just use nullify_over normally.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/hppa/translate.c | 73 +++++++++++------------------------------
 1 file changed, 20 insertions(+), 53 deletions(-)

diff --git a/target/hppa/translate.c b/target/hppa/translate.c
index cbf78a4007..ceba7a98e5 100644
--- a/target/hppa/translate.c
+++ b/target/hppa/translate.c
@@ -1870,17 +1870,15 @@ static bool do_cbranch(DisasContext *ctx, int64_t disp, bool is_n,
 static bool do_ibranch(DisasContext *ctx, TCGv_i64 dest,
                        unsigned link, bool is_n)
 {
-    TCGv_i64 a0, a1, next, tmp;
-    TCGCond c;
+    TCGv_i64 next;
 
-    assert(ctx->null_lab == NULL);
+    if (ctx->null_cond.c == TCG_COND_NEVER && ctx->null_lab == NULL) {
+        next = tcg_temp_new_i64();
+        tcg_gen_mov_i64(next, dest);
 
-    if (ctx->null_cond.c == TCG_COND_NEVER) {
         if (link != 0) {
             copy_iaoq_entry(ctx, cpu_gr[link], ctx->iaoq_n, ctx->iaoq_n_var);
         }
-        next = tcg_temp_new_i64();
-        tcg_gen_mov_i64(next, dest);
         if (is_n) {
             if (use_nullify_skip(ctx)) {
                 copy_iaoq_entry(ctx, cpu_iaoq_f, -1, next);
@@ -1894,60 +1892,29 @@ static bool do_ibranch(DisasContext *ctx, TCGv_i64 dest,
         }
         ctx->iaoq_n = -1;
         ctx->iaoq_n_var = next;
-    } else if (is_n && use_nullify_skip(ctx)) {
-        /* The (conditional) branch, B, nullifies the next insn, N,
-           and we're allowed to skip execution N (no single-step or
-           tracepoint in effect).  Since the goto_ptr that we must use
-           for the indirect branch consumes no special resources, we
-           can (conditionally) skip B and continue execution.  */
-        /* The use_nullify_skip test implies we have a known control path.  */
-        tcg_debug_assert(ctx->iaoq_b != -1);
-        tcg_debug_assert(ctx->iaoq_n != -1);
+        return true;
+    }
 
-        /* We do have to handle the non-local temporary, DEST, before
-           branching.  Since IOAQ_F is not really live at this point, we
-           can simply store DEST optimistically.  Similarly with IAOQ_B.  */
+    nullify_over(ctx);
+
+    if (is_n && use_nullify_skip(ctx)) {
         copy_iaoq_entry(ctx, cpu_iaoq_f, -1, dest);
         next = tcg_temp_new_i64();
         tcg_gen_addi_i64(next, dest, 4);
         copy_iaoq_entry(ctx, cpu_iaoq_b, -1, next);
-
-        nullify_over(ctx);
-        if (link != 0) {
-            copy_iaoq_entry(ctx, cpu_gr[link], ctx->iaoq_n, ctx->iaoq_n_var);
-        }
-        tcg_gen_lookup_and_goto_ptr();
-        return nullify_end(ctx);
+        nullify_set(ctx, 0);
     } else {
-        c = ctx->null_cond.c;
-        a0 = ctx->null_cond.a0;
-        a1 = ctx->null_cond.a1;
-
-        tmp = tcg_temp_new_i64();
-        next = tcg_temp_new_i64();
-
-        copy_iaoq_entry(ctx, tmp, ctx->iaoq_n, ctx->iaoq_n_var);
-        tcg_gen_movcond_i64(c, next, a0, a1, tmp, dest);
-        ctx->iaoq_n = -1;
-        ctx->iaoq_n_var = next;
-
-        if (link != 0) {
-            tcg_gen_movcond_i64(c, cpu_gr[link], a0, a1, cpu_gr[link], tmp);
-        }
-
-        if (is_n) {
-            /* The branch nullifies the next insn, which means the state of N
-               after the branch is the inverse of the state of N that applied
-               to the branch.  */
-            tcg_gen_setcond_i64(tcg_invert_cond(c), cpu_psw_n, a0, a1);
-            cond_free(&ctx->null_cond);
-            ctx->null_cond = cond_make_n();
-            ctx->psw_n_nonzero = true;
-        } else {
-            cond_free(&ctx->null_cond);
-        }
+        copy_iaoq_entry(ctx, cpu_iaoq_f, ctx->iaoq_b, cpu_iaoq_b);
+        copy_iaoq_entry(ctx, cpu_iaoq_b, -1, dest);
+        nullify_set(ctx, is_n);
     }
-    return true;
+    if (link != 0) {
+        copy_iaoq_entry(ctx, cpu_gr[link], ctx->iaoq_n, ctx->iaoq_n_var);
+    }
+
+    tcg_gen_lookup_and_goto_ptr();
+    ctx->base.is_jmp = DISAS_NORETURN;
+    return nullify_end(ctx);
 }
 
 /* Implement
-- 
2.34.1



  parent reply	other threads:[~2024-04-25  0:12 UTC|newest]

Thread overview: 50+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-04-24 23:59 [PATCH 00/45] target/hppa: Misc improvements Richard Henderson
2024-04-24 23:59 ` [PATCH 01/45] target/hppa: Move cpu_get_tb_cpu_state out of line Richard Henderson
2024-04-24 23:59 ` [PATCH 02/45] target/hppa: Use hppa_form_gva_psw in hppa_cpu_get_pc Richard Henderson
2024-04-24 23:59 ` [PATCH 03/45] target/hppa: Move constant destination check into use_goto_tb Richard Henderson
2024-04-24 23:59 ` [PATCH 04/45] target/hppa: Pass displacement to do_dbranch Richard Henderson
2024-05-13 11:41   ` Philippe Mathieu-Daudé
2024-04-24 23:59 ` Richard Henderson [this message]
2024-04-24 23:59 ` [PATCH 06/45] target/hppa: Use CF_BP_PAGE instead of cpu_breakpoint_test Richard Henderson
2024-04-24 23:59 ` [PATCH 07/45] target/hppa: Add install_iaq_entries Richard Henderson
2024-04-24 23:59 ` [PATCH 08/45] target/hppa: Add install_link Richard Henderson
2024-04-24 23:59 ` [PATCH 09/45] target/hppa: Delay computation of IAQ_Next Richard Henderson
2024-04-24 23:59 ` [PATCH 10/45] target/hppa: Skip nullified insns in unconditional dbranch path Richard Henderson
2024-04-24 23:59 ` [PATCH 11/45] target/hppa: Simplify TB end Richard Henderson
2024-04-24 23:59 ` [PATCH 12/45] target/hppa: Add IASQ entries to DisasContext Richard Henderson
2024-04-24 23:59 ` [PATCH 13/45] target/hppa: Add space arguments to install_iaq_entries Richard Henderson
2024-04-24 23:59 ` [PATCH 14/45] target/hppa: Add space argument to do_ibranch Richard Henderson
2024-04-24 23:59 ` [PATCH 15/45] target/hppa: Use umax in do_ibranch_priv Richard Henderson
2024-04-24 23:59 ` [PATCH 16/45] target/hppa: Always make a copy " Richard Henderson
2024-04-24 23:59 ` [PATCH 17/45] target/hppa: Introduce and use DisasIAQE for branch management Richard Henderson
2024-04-24 23:59 ` [PATCH 18/45] target/hppa: Use displacements in DisasIAQE Richard Henderson
2024-04-24 23:59 ` [PATCH 19/45] target/hppa: Rename cond_make_* helpers Richard Henderson
2024-04-24 23:59 ` [PATCH 20/45] target/hppa: Use TCG_COND_TST* in do_cond Richard Henderson
2024-04-24 23:59 ` [PATCH 21/45] target/hppa: Use TCG_COND_TST* in do_log_cond Richard Henderson
2024-04-25  0:00 ` [PATCH 22/45] target/hppa: Use TCG_COND_TST* in do_unit_zero_cond Richard Henderson
2024-04-25  0:00 ` [PATCH 23/45] target/hppa: Use TCG_COND_TST* in do_unit_addsub Richard Henderson
2024-04-25  0:00 ` [PATCH 24/45] target/hppa: Use TCG_COND_TST* in trans_bb_imm Richard Henderson
2024-04-25  0:00 ` [PATCH 25/45] target/hppa: Use registerfields.h for FPSR Richard Henderson
2024-04-25  0:00 ` [PATCH 26/45] target/hppa: Use TCG_COND_TST* in trans_ftest Richard Henderson
2024-04-25  0:00 ` [PATCH 27/45] target/hppa: Remove cond_free Richard Henderson
2024-04-25  0:00 ` [PATCH 28/45] target/hppa: Introduce DisasDelayException Richard Henderson
2024-04-25  0:00 ` [PATCH 29/45] target/hppa: Use delay_excp for conditional traps Richard Henderson
2024-04-25  0:00 ` [PATCH 30/45] target/hppa: Use delay_excp for conditional trap on overflow Richard Henderson
2024-04-25  0:00 ` [PATCH 31/45] linux-user/hppa: Force all code addresses to PRIV_USER Richard Henderson
2024-04-25  0:00 ` [PATCH 32/45] target/hppa: Store full iaoq_f and page bits of iaoq_d in TB Richard Henderson
2024-04-25  0:00 ` [PATCH 33/45] target/hppa: Do not mask in copy_iaoq_entry Richard Henderson
2024-04-25  0:00 ` [PATCH 34/45] target/hppa: Improve hppa_cpu_dump_state Richard Henderson
2024-04-25  0:00 ` [PATCH 35/45] target/hppa: Split PSW X and B into their own field Richard Henderson
2024-04-25  0:00 ` [PATCH 36/45] target/hppa: Manage PSW_X and PSW_B in translator Richard Henderson
2024-04-25  0:00 ` [PATCH 37/45] target/hppa: Implement PSW_B Richard Henderson
2024-04-25  0:00 ` [PATCH 38/45] target/hppa: Implement PSW_X Richard Henderson
2024-04-25  0:00 ` [PATCH 39/45] target/hppa: Drop tlb_entry return from hppa_get_physical_address Richard Henderson
2024-04-25  0:00 ` [PATCH 40/45] target/hppa: Adjust priv for B,GATE at runtime Richard Henderson
2024-04-25  0:00 ` [PATCH 41/45] target/hppa: Implement CF_PCREL Richard Henderson
2024-04-25  0:00 ` [PATCH 42/45] target/hppa: Implement PSW_T Richard Henderson
2024-04-25  0:00 ` [PATCH 43/45] target/hppa: Implement PSW_H, PSW_L Richard Henderson
2024-04-25  0:00 ` [PATCH 44/45] target/hppa: Log cpu state at interrupt Richard Henderson
2024-04-25  0:00 ` [PATCH 45/45] target/hppa: Log cpu state on return-from-interrupt Richard Henderson
2024-05-10 14:48 ` [PATCH 00/45] target/hppa: Misc improvements Philippe Mathieu-Daudé
2024-05-12 16:08   ` Sven Schnelle
2024-05-13  6:11     ` Helge Deller

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20240425000023.1002026-6-richard.henderson@linaro.org \
    --to=richard.henderson@linaro.org \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).