[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: /srv/bzr/emacs/trunk r103781: * src/bytecode.c (Fbyte_code): CAR and
From: |
Daniel Colascione |
Subject: |
Re: /srv/bzr/emacs/trunk r103781: * src/bytecode.c (Fbyte_code): CAR and CDR can GC. |
Date: |
Wed, 30 Mar 2011 12:41:56 -0700 |
User-agent: |
Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.2.15) Gecko/20110303 Thunderbird/3.1.9 |
On 3/30/2011 11:04 AM, Stefan Monnier wrote:
------------------------------------------------------------
revno: 103781
committer: Stefan Monnier<address@hidden>
branch nick: trunk
timestamp: Wed 2011-03-30 14:04:11 -0400
message:
* src/bytecode.c (Fbyte_code): CAR and CDR can GC.
Wait, what? How? wrong_type_argument never returns. If we run into a
problem, we'll throw up to some higher context and never hit the
AFTER_POTENTIAL_GC case. BEFORE_POTENTIAL_GC (if it's not a noop) sets
the stack top field to point to the top of the stack, but because we
never return from wrong_type_argument, we don't care whether the stack
is lost. v1 might be on the stack, so the only thing keeping it alive
might be stack.top.
Now, Fsignal *does* run signal-hook-function, which of course might GC.
But that's done using call2, which will keep error_symbol and data
alive even if we do GC, so a GC in this case shouldn't actually cause
any problems. Likewise for the debugging hooks. So GC shouldn't
actually cause a problem, IIUC.
What am I missing?
=== modified file 'src/ChangeLog'
--- a/src/ChangeLog 2011-03-30 13:35:37 +0000
+++ b/src/ChangeLog 2011-03-30 18:04:11 +0000
@@ -1,3 +1,7 @@
+2011-03-30 Stefan Monnier <address@hidden>
+
+ * bytecode.c (Fbyte_code): CAR and CDR can GC.
+
2011-03-30 Zachary Kanfer <address@hidden> (tiny change)
* keyboard.c (Fexecute_extended_command): Do log the "suggest key
=== modified file 'src/bytecode.c'
--- a/src/bytecode.c 2011-03-17 02:18:00 +0000
+++ b/src/bytecode.c 2011-03-30 18:04:11 +0000
@@ -554,7 +554,16 @@
{
Lisp_Object v1;
v1 = TOP;
- TOP = CAR (v1);
+ if (CONSP (v1))
+ TOP = XCAR (v1);
+ else if (NILP (v1))
+ TOP = Qnil;
+ else
+ {
+ BEFORE_POTENTIAL_GC ();
+ wrong_type_argument (Qlistp, v1);
+ AFTER_POTENTIAL_GC ();
+ }
break;
}
@@ -580,7 +589,17 @@
{
Lisp_Object v1;
v1 = TOP;
- TOP = CDR (v1);
+ if (CONSP (v1))
+ TOP = XCDR (v1);
+ else if (NILP (v1))
+ TOP = Qnil;
+ else
+ {
+ BEFORE_POTENTIAL_GC ();
+ wrong_type_argument (Qlistp, v1);
+ AFTER_POTENTIAL_GC ();
+ }
+ break;
break;
}
@@ -911,13 +930,13 @@
v1 = POP;
v2 = TOP;
CHECK_NUMBER (v2);
- AFTER_POTENTIAL_GC ();
op = XINT (v2);
immediate_quit = 1;
while (--op >= 0 && CONSP (v1))
v1 = XCDR (v1);
immediate_quit = 0;
TOP = CAR (v1);
+ AFTER_POTENTIAL_GC ();
break;
}
- Re: /srv/bzr/emacs/trunk r103781: * src/bytecode.c (Fbyte_code): CAR and CDR can GC.,
Daniel Colascione <=