All of lore.kernel.org
 help / color / mirror / Atom feed
From: Paolo Bonzini <bonzini@gnu.org>
To: Blue Swirl <blauwirbel@gmail.com>
Cc: qemu-devel@nongnu.org
Subject: [Qemu-devel] Re: [PATCH 0/2] simplify global register save/restore
Date: Sat, 13 Feb 2010 21:26:42 +0100	[thread overview]
Message-ID: <4B770B02.4030409@gnu.org> (raw)
In-Reply-To: <f43fc5581002130958q17edaf68tf3d66c319059b617@mail.gmail.com>

[-- Attachment #1: Type: text/plain, Size: 472 bytes --]

On 02/13/2010 06:58 PM, Blue Swirl wrote:
> V9 ABI gives more registers to application use.

Except that glibc uses those---in theory, as I see it, it should be 
compiled with fixed g2 and g3 to leave them to the application.  I get 
it now.

It may be possible to make the workaround a big less ugly (I'm thinking 
of avoiding #undef/#define by using assembly).  I made a patch (see 
attachment, just FYI), maybe sometime I'll try it using self-virtualized 
qemu.

Paolo

[-- Attachment #2: sparc --]
[-- Type: text/plain, Size: 2730 bytes --]

commit 59ca12838278bed97ce5cc311f90ddfec7953047
Author: Paolo Bonzini <pbonzini@redhat.com>
Date:   Sat Feb 13 21:13:12 2010 +0100

    make sparc workaround less ugly
    
    Not-quite-signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

diff --git a/cpu-exec.c b/cpu-exec.c
index badd5d7..01b7143 100644
--- a/cpu-exec.c
+++ b/cpu-exec.c
@@ -39,9 +39,14 @@
 #endif
 
 #if defined(__sparc__) && !defined(CONFIG_SOLARIS)
-// Work around ugly bugs in glibc that mangle global register contents
-#undef env
+/* glibc will mangle global register contents.  To work around this,
+ * we avoid using the global register in this file, and place back
+ * cpu_single_env in AREG0 before giving control to target-* routines.
+ */
+#define export_env()	asm ("mov %0, %%" AREG0 : : "r" (cpu_single_env) : AREG0); 
 #define env cpu_single_env
+#else
+#define export_env()
 #endif
 
 int tb_invalidated_flag;
@@ -257,11 +262,7 @@ int cpu_exec(CPUState *env1)
     /* prepare setjmp context for exception handling */
     for(;;) {
         if (setjmp(env->jmp_env) == 0) {
-#if defined(__sparc__) && !defined(CONFIG_SOLARIS)
-#undef env
-                    env = cpu_single_env;
-#define env cpu_single_env
-#endif
+            export_env();
             /* if an exception is pending, we execute it here */
             if (env->exception_index >= 0) {
                 if (env->exception_index >= EXCP_INTERRUPT) {
@@ -387,11 +388,7 @@ int cpu_exec(CPUState *env1)
                             env->interrupt_request &= ~(CPU_INTERRUPT_HARD | CPU_INTERRUPT_VIRQ);
                             intno = cpu_get_pic_interrupt(env);
                             qemu_log_mask(CPU_LOG_TB_IN_ASM, "Servicing hardware INT=0x%02x\n", intno);
-#if defined(__sparc__) && !defined(CONFIG_SOLARIS)
-#undef env
-                    env = cpu_single_env;
-#define env cpu_single_env
-#endif
+                            export_env();
                             do_interrupt(intno, 0, 0, 0, 1);
                             /* ensure that no TB jump will be modified as
                                the program flow was changed */
@@ -603,12 +600,8 @@ int cpu_exec(CPUState *env1)
                 if (!unlikely (env->exit_request)) {
                     env->current_tb = tb;
                     tc_ptr = tb->tc_ptr;
-                /* execute the generated code */
-#if defined(__sparc__) && !defined(CONFIG_SOLARIS)
-#undef env
-                    env = cpu_single_env;
-#define env cpu_single_env
-#endif
+                    /* execute the generated code */
+                    export_env();
                     next_tb = tcg_qemu_tb_exec(tc_ptr);
                     env->current_tb = NULL;
                     if ((next_tb & 3) == 2) {

  reply	other threads:[~2010-02-13 20:26 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-02-10 23:26 [Qemu-devel] [PATCH 0/2] simplify global register save/restore Paolo Bonzini
2010-02-10 23:26 ` [Qemu-devel] [PATCH 1/2] remove dead m68k global register definitions Paolo Bonzini
2010-02-10 23:26 ` [Qemu-devel] [PATCH 2/2] get rid of hostregs_helper.h Paolo Bonzini
2010-02-18 19:07   ` Blue Swirl
2010-02-18 20:25     ` [Qemu-devel] [PATCH v2 1/2] remove dead m68k definitions Paolo Bonzini
2010-02-18 20:25       ` [Qemu-devel] [PATCH v2 2/2] get rid of hostregs_helper.h Paolo Bonzini
2010-02-18 21:28         ` [Qemu-devel] " Blue Swirl
2010-02-25 11:40           ` Michael S. Tsirkin
2010-02-25 12:50             ` Paolo Bonzini
2010-02-25 13:04               ` Michael S. Tsirkin
2010-02-25 13:11                 ` Paolo Bonzini
2010-02-13 17:58 ` [Qemu-devel] [PATCH 0/2] simplify global register save/restore Blue Swirl
2010-02-13 20:26   ` Paolo Bonzini [this message]
2010-02-13 20:57     ` [Qemu-devel] " Blue Swirl
2010-02-13 20:58       ` Paolo Bonzini
2010-02-26 11:30 ` [Qemu-devel] " Paul Brook
2010-02-26 13:05   ` Paolo Bonzini
2010-02-26 18:32     ` Paul Brook
2010-03-01 12:30       ` Paolo Bonzini
2010-03-01 14:02         ` Paul Brook

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=4B770B02.4030409@gnu.org \
    --to=bonzini@gnu.org \
    --cc=blauwirbel@gmail.com \
    --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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.