qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Richard Henderson <rth@twiddle.net>
To: qemu-devel@nongnu.org
Cc: Ulrich Weigand <ulrich.weigand@de.ibm.com>, av1474@comtv.ru
Subject: [Qemu-devel] [PATCH 12/26] tcg-ppc64: Adjust tcg_out_call for ELFv2
Date: Thu,  1 May 2014 08:44:33 -0700	[thread overview]
Message-ID: <1398959087-23590-13-git-send-email-rth@twiddle.net> (raw)
In-Reply-To: <1398959087-23590-1-git-send-email-rth@twiddle.net>

From: Ulrich Weigand <ulrich.weigand@de.ibm.com>

The new ELFv2 ABI, used by default on powerpc64le-linux hosts,
introduced some changes that are incompatible with code currently
generated by the ppc64 TGC target.  In particular, we no longer
use function descriptors.

This patch adds support for the ELFv2 ABI in the ppc64 TGC
function call and function prologue sequences.

Signed-off-by: Ulrich Weigand <ulrich.weigand@de.ibm.com>
Signed-off-by: Richard Henderson <rth@twiddle.net>
---
 tcg/ppc64/tcg-target.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

diff --git a/tcg/ppc64/tcg-target.c b/tcg/ppc64/tcg-target.c
index d3cc237..635ff98 100644
--- a/tcg/ppc64/tcg-target.c
+++ b/tcg/ppc64/tcg-target.c
@@ -1115,6 +1115,23 @@ static void tcg_out_call(TCGContext *s, tcg_insn_unit *target)
         tcg_out_ld(s, TCG_TYPE_PTR, TCG_REG_R2, TCG_REG_R2, ofs + SZP);
         tcg_out32(s, BCCTR | BO_ALWAYS | LK);
     }
+#elif defined(_CALL_ELF) && _CALL_ELF == 2
+    intptr_t diff;
+
+    /* In the ELFv2 ABI, we have to set up r12 to contain the destination
+       address, which the callee uses to compute its TOC address.  */
+    /* FIXME: when the branch is in range, we could avoid r12 load if we
+       knew that the destination uses the same TOC, and what its local
+       entry point offset is.  */
+    tcg_out_movi(s, TCG_TYPE_PTR, TCG_REG_R12, (intptr_t)target);
+
+    diff = tcg_pcrel_diff(s, target);
+    if (in_range_b(diff)) {
+        tcg_out_b(s, LK, target);
+    } else {
+        tcg_out32(s, MTSPR | RS(TCG_REG_R12) | CTR);
+        tcg_out32(s, BCCTR | BO_ALWAYS | LK);
+    }
 #else
     tcg_out_b(s, LK, target);
 #endif
-- 
1.9.0

  parent reply	other threads:[~2014-05-01 15:45 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-05-01 15:44 [Qemu-devel] [PATCH 00/26] Merge ppc32/ppc64 tcg backends Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 01/26] tcg-ppc: Use uintptr_t in ppc_tb_set_jmp_target Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 02/26] tcg-ppc64: Avoid some hard-codings of TCG_TYPE_I64 Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 03/26] tcg-ppc64: Move functions around Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 04/26] tcg-ppc64: Relax register restrictions in tcg_out_mem_long Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 05/26] tcg-ppc64: Use tcg_out_{ld, st, cmp} internally Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 06/26] tcg-ppc64: Make TCG_AREG0 and TCG_REG_CALL_STACK enum constants Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 07/26] tcg-ppc64: Move call macros out of tcg-target.h Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 08/26] tcg-ppc64: Fix TCG_TARGET_CALL_STACK_OFFSET Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 09/26] tcg-ppc64: Better parameterize the stack frame Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 10/26] tcg-ppc64: Use the correct test in tcg_out_call Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 11/26] tcg-ppc64: Support the ppc64 elfv2 ABI Richard Henderson
2014-05-01 15:44 ` Richard Henderson [this message]
2014-05-01 15:44 ` [Qemu-devel] [PATCH 13/26] tcg-ppc64: Merge 32-bit ABIs into the prologue / frame code Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 14/26] tcg-ppc64: Fix sub2 implementation Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 15/26] tcg-ppc64: Begin merging ppc32 with ppc64 Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 16/26] tcg-ppc64: Merge ppc32 brcond2, setcond2, muluh Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 17/26] tcg-ppc64: Merge ppc32 qemu_ld/st Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 18/26] tcg-ppc64: Merge ppc32 register usage Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 19/26] tcg-ppc64: Support mulsh_i32 Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 20/26] tcg-ppc64: Merge ppc32 shifts Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 21/26] tcg-ppc: Remove the backend Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 22/26] tcg-ppc: Rename the tcg/ppc64 backend Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 23/26] qemu/osdep: Remove the need for qemu_init_auxval Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 24/26] tcg-ppc: Merge cache-utils into the backend Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 25/26] tcg-ppc64: Use the return address as a base pointer Richard Henderson
2014-05-01 15:44 ` [Qemu-devel] [PATCH 26/26] tcg-ppc: Streamline USE_DIRECT_JUMP Richard Henderson
2014-05-02 14:56 ` [Qemu-devel] [PATCH 00/26] Merge ppc32/ppc64 tcg backends Tom Musta
2014-05-02 16:30 ` Ulrich Weigand
2014-05-02 16:43   ` Richard Henderson
2014-05-05 20:32     ` Tom Musta

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=1398959087-23590-13-git-send-email-rth@twiddle.net \
    --to=rth@twiddle.net \
    --cc=av1474@comtv.ru \
    --cc=qemu-devel@nongnu.org \
    --cc=ulrich.weigand@de.ibm.com \
    /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).