summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorFather Chrysostomos <sprout@cpan.org>2011-08-14 19:16:14 -0700
committerFather Chrysostomos <sprout@cpan.org>2011-08-15 06:20:19 -0700
commitf79aa60b66082c8bff80f325979742bfb6c73709 (patch)
tree9abac236ef8e21df8d6080190f3e12e2aae9eadd
parent186abd568524c217f6dbd29891eef21c3cd2a3b0 (diff)
downloadperl-f79aa60b66082c8bff80f325979742bfb6c73709.tar.gz
Make lock(&foo) syntax nominally lock the subroutine
In 5.10, lock(&foo) was an error for non-lvalue subs. For lvalue subs, it passed &foo to the lockhook and return \&foo. In 5.12, lock(&foo) was still an error for non-lvalue subs. For lvalue subs, it would pass &foo to the lockhook and then either trip an assertion (-DDEBUGGING) or return &foo, resulting in inter- esting bugs. Commit f4df43b5e changed lock(&lvalue_sub) to call the sub and lock its return value. As Reini Urban pointed out in <CAHiT=DE5cVZbuCR3kb=Q5oCa18vo3jr5jZKmURHYha2PwF4pEQ@mail.gmail.com>, locking a subroutine does have its uses. Since lock(&foo) has never really worked anyway, we can still change this. So, for lvalue subs, this reverts back to the 5.10 behaviour. For non-lvalue subs, it now behaves the same way, the lvalue flag making no difference. Note that it still causes an error at run-time, if threads::shared is loaded, as its lockhook is conservative in what it accepts. But this change allows for future extensibility, unlike f4df43b5e. A note about the implementation: There are two pieces of code (at least) in op.c that convert an entersub op into an rv2cv, one in S_doref and the other in Perl_op_lvalue_flags. Originally (before f4df43b5e) it was S_doref that took care of that for OP_LOCK. But Perl_op_lvalue_flags is called first, so it would assume it was an assignment to a sub call and croak if there was no lvalue sub in the symbol table. This commit adds back the special case for OP_LOCK, but in Perl_op_lvalue_flags, not S_doref.
-rw-r--r--op.c3
-rw-r--r--pod/perldelta.pod13
-rw-r--r--pod/perlfunc.pod2
-rw-r--r--pp.c4
-rw-r--r--t/op/cproto.t2
-rw-r--r--t/op/lock.t6
6 files changed, 16 insertions, 14 deletions
diff --git a/op.c b/op.c
index 4f8693a4b9..b3c5f86f31 100644
--- a/op.c
+++ b/op.c
@@ -1705,7 +1705,7 @@ Perl_op_lvalue_flags(pTHX_ OP *o, I32 type, U32 flags)
break;
goto nomod;
case OP_ENTERSUB:
- if ((type == OP_UNDEF || type == OP_REFGEN) &&
+ if ((type == OP_UNDEF || type == OP_REFGEN || type == OP_LOCK) &&
!(o->op_flags & OPf_STACKED)) {
o->op_type = OP_RV2CV; /* entersub => rv2cv */
/* Both ENTERSUB and RV2CV use this bit, but for different pur-
@@ -10415,6 +10415,7 @@ Perl_core_prototype(pTHX_ SV *sv, const char *name, const int code,
str[n++] = '$';
str[n++] = '@';
str[n++] = '%';
+ if (i == OP_LOCK) str[n++] = '&';
str[n++] = '*';
str[n++] = ']';
}
diff --git a/pod/perldelta.pod b/pod/perldelta.pod
index 0d87c3aea0..2309a09ad7 100644
--- a/pod/perldelta.pod
+++ b/pod/perldelta.pod
@@ -480,14 +480,17 @@ L</Modules and Pragmata>.
=item *
-Locking an lvalue subroutine (via C<lock &lvsub>) now locks the return
-value, instead of trying to lock the sub (which has no effect). It also no
-longer tries to return the sub as a scalar, resulting in strange side
-effects like C<ref \$_> returning "CODE" in some instances.
+Locking a subroutine (via C<lock &sub>) is no longer a compile-time error
+for regular subs. For lvalue subroutines, it no longer tries to return the
+sub as a scalar, resulting in strange side effects like C<ref \$_>
+returning "CODE" in some instances.
+
+C<lock &sub> is now a run-time error if L<threads::shared> is loaded (a
+no-op otherwise), but that may be rectified in a future version.
=item *
-C<lock>'s prototype has been corrected to C<(\[$@%*])> from C<(\$)>, which
+C<lock>'s prototype has been corrected to C<(\[$@%&*])> from C<(\$)>, which
was just wrong.
=item *
diff --git a/pod/perlfunc.pod b/pod/perlfunc.pod
index 264fbdcbcd..04c6a05bf9 100644
--- a/pod/perlfunc.pod
+++ b/pod/perlfunc.pod
@@ -3095,7 +3095,7 @@ This function places an advisory lock on a shared variable or referenced
object contained in I<THING> until the lock goes out of scope.
The value returned is the scalar itself, if the argument is a scalar, or a
-reference, if the argument is a hash or array.
+reference, if the argument is a hash, array or subroutine.
lock() is a "weak keyword" : this means that if you've defined a function
by this name (before any calls to it), that function will be called
diff --git a/pp.c b/pp.c
index ca94935c30..3421d9701d 100644
--- a/pp.c
+++ b/pp.c
@@ -5914,9 +5914,9 @@ PP(pp_lock)
dSP;
dTOPss;
SV *retsv = sv;
- assert(SvTYPE(retsv) != SVt_PVCV);
SvLOCK(sv);
- if (SvTYPE(retsv) == SVt_PVAV || SvTYPE(retsv) == SVt_PVHV) {
+ if (SvTYPE(retsv) == SVt_PVAV || SvTYPE(retsv) == SVt_PVHV
+ || SvTYPE(retsv) == SVt_PVCV) {
retsv = refto(retsv);
}
SETs(retsv);
diff --git a/t/op/cproto.t b/t/op/cproto.t
index 1476ae6ab4..e995416fb3 100644
--- a/t/op/cproto.t
+++ b/t/op/cproto.t
@@ -138,7 +138,7 @@ link ($$)
listen (*$)
local undef
localtime (;$)
-lock (\[$@%*])
+lock (\[$@%&*])
log (_)
lstat (;*)
lt undef
diff --git a/t/op/lock.t b/t/op/lock.t
index 2fd67821eb..c40ec4c9fd 100644
--- a/t/op/lock.t
+++ b/t/op/lock.t
@@ -10,9 +10,7 @@ plan tests => 5;
is \lock $foo, \$foo, 'lock returns a scalar argument';
is lock @foo, \@foo, 'lock returns a ref to its array argument';
is lock %foo, \%foo, 'lock returns a ref to its hash argument';
-eval { lock &foo }; my $file = __FILE__; my $line = __LINE__;
-is $@, "Can't modify non-lvalue subroutine call at $file line $line.\n",
- 'Error when locking non-lvalue sub';
+is lock &foo, \&foo, 'lock returns a ref to its code argument';
sub eulavl : lvalue { $x }
-is \lock &eulavl, \$x, 'locking lvalue sub acts on retval, just like tie';
+is lock &eulavl, \&eulavl, 'lock returns a ref to its lvalue sub arg';