* [PATCH] Do not propagate self-dependent value (PR ipa/93763)
@ 2020-02-18 15:12 Feng Xue OS
2020-02-21 17:59 ` Martin Jambor
0 siblings, 1 reply; 6+ messages in thread
From: Feng Xue OS @ 2020-02-18 15:12 UTC (permalink / raw)
To: mjambor, Jan Hubicka, gcc-patches
[-- Attachment #1: Type: text/plain, Size: 484 bytes --]
Currently, for self-recursive call, we never use value originated from non-passthrough
jump function as source to avoid propagation explosion, but self-dependent value is
missed. This patch is made to fix the bug.
Bootstrapped/regtested on x86_64-linux and aarch64-linux.
Feng
---
2020-02-18 Feng Xue <fxue@os.amperecomputing.com>
PR ipa/93763
* ipa-cp.c (self_recursively_generated_p): Mark self-dependent value as
self-recursively generated.
[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0001-Do-not-propagate-self-dependent-value-PR-ipa-93763.patch --]
[-- Type: text/x-patch; name="0001-Do-not-propagate-self-dependent-value-PR-ipa-93763.patch", Size: 4217 bytes --]
From 1ff803f33de0fe86d526deb23af2d08c15028ff9 Mon Sep 17 00:00:00 2001
From: Feng Xue <fxue@os.amperecomputing.com>
Date: Mon, 17 Feb 2020 17:07:04 +0800
Subject: [PATCH] Do not propagate self-dependent value (PR ipa/93763)
---
gcc/ipa-cp.c | 10 ++++---
gcc/testsuite/g++.dg/ipa/pr93763.C | 15 ++++++++++
gcc/testsuite/gcc.dg/ipa/pr93763.c | 46 ++++++++++++++++++++++++++++++
3 files changed, 67 insertions(+), 4 deletions(-)
create mode 100644 gcc/testsuite/g++.dg/ipa/pr93763.C
create mode 100644 gcc/testsuite/gcc.dg/ipa/pr93763.c
diff --git a/gcc/ipa-cp.c b/gcc/ipa-cp.c
index 4f5b72e6994..1d0c1ac0f35 100644
--- a/gcc/ipa-cp.c
+++ b/gcc/ipa-cp.c
@@ -1897,8 +1897,8 @@ ipcp_lattice<valtype>::add_value (valtype newval, cgraph_edge *cs,
}
/* Return true, if a ipcp_value VAL is orginated from parameter value of
- self-feeding recursive function by applying non-passthrough arithmetic
- transformation. */
+ self-feeding recursive function via some kind of pass-through jump
+ function. */
static bool
self_recursively_generated_p (ipcp_value<tree> *val)
@@ -1909,10 +1909,12 @@ self_recursively_generated_p (ipcp_value<tree> *val)
{
cgraph_edge *cs = src->cs;
- if (!src->val || cs->caller != cs->callee->function_symbol ()
- || src->val == val)
+ if (!src->val || cs->caller != cs->callee->function_symbol ())
return false;
+ if (src->val == val)
+ continue;
+
if (!info)
info = IPA_NODE_REF (cs->caller);
diff --git a/gcc/testsuite/g++.dg/ipa/pr93763.C b/gcc/testsuite/g++.dg/ipa/pr93763.C
new file mode 100644
index 00000000000..61117108ecd
--- /dev/null
+++ b/gcc/testsuite/g++.dg/ipa/pr93763.C
@@ -0,0 +1,15 @@
+/* { dg-do compile } */
+/* { dg-options "-O3" } */
+
+struct search_param {
+ int total;
+};
+void search_trivial(search_param error_left) {
+ search_trivial(error_left);
+ search_param error_left2{error_left};
+ error_left2.total--;
+ search_trivial(error_left2);
+}
+void search_algo_uni(search_param error_left) { search_trivial(error_left); }
+void search_algo(search_param error_left) { search_algo_uni(error_left); }
+int main() { search_algo({}); return 0; }
diff --git a/gcc/testsuite/gcc.dg/ipa/pr93763.c b/gcc/testsuite/gcc.dg/ipa/pr93763.c
new file mode 100644
index 00000000000..d11705932b6
--- /dev/null
+++ b/gcc/testsuite/gcc.dg/ipa/pr93763.c
@@ -0,0 +1,46 @@
+/* { dg-do compile } */
+/* { dg-options "-O3" } */
+
+typedef struct a a;
+struct a {
+ a *b
+} d;
+e, k, ah, al;
+f(aa) {
+ if (aa & 1)
+ goto g;
+ f(aa | 2);
+g:
+ h();
+}
+l() {
+ {
+ f(072);
+ i(e, d, 92);
+ }
+}
+ag() {
+ { i(e, d, 36); }
+}
+ai(a *m, a *n, unsigned aa) {
+ f(aa);
+ j(k, l, ah, 1);
+}
+j(int c, a m, int aj, int aa) {
+ int ak = aa;
+ { i(e, d, ak); }
+}
+i(int c, a *m, unsigned aa) {
+ {
+ { i(c, (*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(
+*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(
+*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*m).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b)
+.b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b)
+.b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b)
+.b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b)
+.b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b, 0);
+ }
+ }
+ int am = aa;
+ ai(ag, al, am);
+}
--
2.17.1
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] Do not propagate self-dependent value (PR ipa/93763)
2020-02-18 15:12 [PATCH] Do not propagate self-dependent value (PR ipa/93763) Feng Xue OS
@ 2020-02-21 17:59 ` Martin Jambor
2020-02-21 20:23 ` Jeff Law
0 siblings, 1 reply; 6+ messages in thread
From: Martin Jambor @ 2020-02-21 17:59 UTC (permalink / raw)
To: Feng Xue OS, Jan Hubicka, gcc-patches
Hi,
On Tue, Feb 18 2020, Feng Xue OS wrote:
> Currently, for self-recursive call, we never use value originated from non-passthrough
> jump function as source to avoid propagation explosion, but self-dependent value is
> missed. This patch is made to fix the bug.
>
> Bootstrapped/regtested on x86_64-linux and aarch64-linux.
>
> Feng
> ---
> 2020-02-18 Feng Xue <fxue@os.amperecomputing.com>
>
> PR ipa/93763
> * ipa-cp.c (self_recursively_generated_p): Mark self-dependent value as
> self-recursively generated.
Honza, this is OK.
Thanks,
Martin
> From 1ff803f33de0fe86d526deb23af2d08c15028ff9 Mon Sep 17 00:00:00 2001
> From: Feng Xue <fxue@os.amperecomputing.com>
> Date: Mon, 17 Feb 2020 17:07:04 +0800
> Subject: [PATCH] Do not propagate self-dependent value (PR ipa/93763)
>
> ---
> gcc/ipa-cp.c | 10 ++++---
> gcc/testsuite/g++.dg/ipa/pr93763.C | 15 ++++++++++
> gcc/testsuite/gcc.dg/ipa/pr93763.c | 46 ++++++++++++++++++++++++++++++
> 3 files changed, 67 insertions(+), 4 deletions(-)
> create mode 100644 gcc/testsuite/g++.dg/ipa/pr93763.C
> create mode 100644 gcc/testsuite/gcc.dg/ipa/pr93763.c
>
> diff --git a/gcc/ipa-cp.c b/gcc/ipa-cp.c
> index 4f5b72e6994..1d0c1ac0f35 100644
> --- a/gcc/ipa-cp.c
> +++ b/gcc/ipa-cp.c
> @@ -1897,8 +1897,8 @@ ipcp_lattice<valtype>::add_value (valtype newval, cgraph_edge *cs,
> }
>
> /* Return true, if a ipcp_value VAL is orginated from parameter value of
> - self-feeding recursive function by applying non-passthrough arithmetic
> - transformation. */
> + self-feeding recursive function via some kind of pass-through jump
> + function. */
>
> static bool
> self_recursively_generated_p (ipcp_value<tree> *val)
> @@ -1909,10 +1909,12 @@ self_recursively_generated_p (ipcp_value<tree> *val)
> {
> cgraph_edge *cs = src->cs;
>
> - if (!src->val || cs->caller != cs->callee->function_symbol ()
> - || src->val == val)
> + if (!src->val || cs->caller != cs->callee->function_symbol ())
> return false;
>
> + if (src->val == val)
> + continue;
> +
> if (!info)
> info = IPA_NODE_REF (cs->caller);
>
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] Do not propagate self-dependent value (PR ipa/93763)
2020-02-21 17:59 ` Martin Jambor
@ 2020-02-21 20:23 ` Jeff Law
2020-02-24 10:10 ` Christophe Lyon
0 siblings, 1 reply; 6+ messages in thread
From: Jeff Law @ 2020-02-21 20:23 UTC (permalink / raw)
To: Martin Jambor, Feng Xue OS, Jan Hubicka, gcc-patches
On Fri, 2020-02-21 at 18:59 +0100, Martin Jambor wrote:
> Hi,
>
> On Tue, Feb 18 2020, Feng Xue OS wrote:
> > Currently, for self-recursive call, we never use value originated from non-passthrough
> > jump function as source to avoid propagation explosion, but self-dependent value is
> > missed. This patch is made to fix the bug.
> >
> > Bootstrapped/regtested on x86_64-linux and aarch64-linux.
> >
> > Feng
> > ---
> > 2020-02-18 Feng Xue <fxue@os.amperecomputing.com>
> >
> > PR ipa/93763
> > * ipa-cp.c (self_recursively_generated_p): Mark self-dependent value as
> > self-recursively generated.
>
> Honza, this is OK.
Thanks. I went ahead and installed Feng's patch.
jeff
>
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] Do not propagate self-dependent value (PR ipa/93763)
2020-02-21 20:23 ` Jeff Law
@ 2020-02-24 10:10 ` Christophe Lyon
2020-02-24 11:50 ` [PATCH] testsuite: Fix recently added ipa testcases [PR93763] Jakub Jelinek
0 siblings, 1 reply; 6+ messages in thread
From: Christophe Lyon @ 2020-02-24 10:10 UTC (permalink / raw)
To: Jeff Law; +Cc: Martin Jambor, Feng Xue OS, Jan Hubicka, gcc-patches
Hi,
On Fri, 21 Feb 2020 at 21:23, Jeff Law <law@redhat.com> wrote:
>
> On Fri, 2020-02-21 at 18:59 +0100, Martin Jambor wrote:
> > Hi,
> >
> > On Tue, Feb 18 2020, Feng Xue OS wrote:
> > > Currently, for self-recursive call, we never use value originated from non-passthrough
> > > jump function as source to avoid propagation explosion, but self-dependent value is
> > > missed. This patch is made to fix the bug.
> > >
> > > Bootstrapped/regtested on x86_64-linux and aarch64-linux.
> > >
> > > Feng
> > > ---
> > > 2020-02-18 Feng Xue <fxue@os.amperecomputing.com>
> > >
> > > PR ipa/93763
> > > * ipa-cp.c (self_recursively_generated_p): Mark self-dependent value as
> > > self-recursively generated.
> >
> > Honza, this is OK.
> Thanks. I went ahead and installed Feng's patch.
>
I'm seeing errors with the new testcase:
end of struct or union
/gcc/testsuite/gcc.dg/ipa/pr93763.c:8:1: warning: data definition has
no type or storage class
/gcc/testsuite/gcc.dg/ipa/pr93763.c:8:1: warning: type defaults to
'int' in declaration of 'e' [-Wimplicit-int]
/gcc/testsuite/gcc.dg/ipa/pr93763.c:8:4: warning: type defaults to
'int' in declaration of 'k' [-Wimplicit-int]
/gcc/testsuite/gcc.dg/ipa/pr93763.c:8:7: warning: type defaults to
'int' in declaration of 'ah' [-Wimplicit-int]
/gcc/testsuite/gcc.dg/ipa/pr93763.c:8:11: warning: type defaults to
'int' in declaration of 'al' [-Wimplicit-int]
/gcc/testsuite/gcc.dg/ipa/pr93763.c:9:1: warning: return type defaults
to 'int' [-Wimplicit-int]
/gcc/testsuite/gcc.dg/ipa/pr93763.c:9:1: warning: type of 'aa'
defaults to 'int' [-Wimplicit-int]
/gcc/testsuite/gcc.dg/ipa/pr93763.c:14:3: warning: implicit
declaration of function 'h' [-Wimplicit-function-declaration]
/gcc/testsuite/gcc.dg/ipa/pr93763.c:16:1: warning: return type
defaults to 'int' [-Wimplicit-int]
/gcc/testsuite/gcc.dg/ipa/pr93763.c:19:5: warning: implicit
declaration of function 'i' [-Wimplicit-function-declaration]
/gcc/testsuite/gcc.dg/ipa/pr93763.c:22:1: warning: return type
defaults to 'int' [-Wimplicit-int]
/gcc/testsuite/gcc.dg/ipa/pr93763.c:25:1: warning: return type
defaults to 'int' [-Wimplicit-int]
/gcc/testsuite/gcc.dg/ipa/pr93763.c:27:3: warning: implicit
declaration of function 'j' [-Wimplicit-function-declaration]
/gcc/testsuite/gcc.dg/ipa/pr93763.c:29:1: warning: return type
defaults to 'int' [-Wimplicit-int]
/gcc/testsuite/gcc.dg/ipa/pr93763.c:33:1: warning: return type
defaults to 'int' [-Wimplicit-int]
/gcc/testsuite/gcc.dg/ipa/pr93763.c:45:6: warning: passing argument 1
of 'ai' from incompatible pointer type [-Wincompatible-pointer-types]
/gcc/testsuite/gcc.dg/ipa/pr93763.c:45:10: warning: passing argument 2 of 'ai'
I thought I could fix what seemed obvious typos, but it seems the
testcase is bogus (missing types, incompatible arguments, ....)
I'm probably missing something, but I'm not the only one according to
gcc-testresults: that test fails on many targets.
Christophe
> jeff
> >
>
^ permalink raw reply [flat|nested] 6+ messages in thread
* [PATCH] testsuite: Fix recently added ipa testcases [PR93763]
2020-02-24 10:10 ` Christophe Lyon
@ 2020-02-24 11:50 ` Jakub Jelinek
2020-02-24 12:33 ` Christophe Lyon
0 siblings, 1 reply; 6+ messages in thread
From: Jakub Jelinek @ 2020-02-24 11:50 UTC (permalink / raw)
To: Jeff Law, Christophe Lyon
Cc: Martin Jambor, Feng Xue OS, Jan Hubicka, gcc-patches
On Mon, Feb 24, 2020 at 11:09:58AM +0100, Christophe Lyon wrote:
> I'm seeing errors with the new testcase:
> end of struct or union
> /gcc/testsuite/gcc.dg/ipa/pr93763.c:8:1: warning: data definition has
> no type or storage class
...
Seems the test has been badly reduced (if the original doesn't emit
warnings, it is always better in the reduction script avoid introducing new
ones).
Also, the g++.dg/ipa/ test fails with -std=c++98 because it is written in
C++11.
Fixed thusly, ok for trunk?
2020-02-24 Jakub Jelinek <jakub@redhat.com>
PR ipa/93763
* gcc.dg/ipa/pr93763.c: Adjust the test so that it compiles without
warnings and still ICEs before the ipa-cp.c fix.
* g++.dg/ipa/pr93763.C: Require c++11 effective target.
--- gcc/testsuite/gcc.dg/ipa/pr93763.c.jj 2020-02-24 12:44:09.626326146 +0100
+++ gcc/testsuite/gcc.dg/ipa/pr93763.c 2020-02-24 12:40:48.707320934 +0100
@@ -3,44 +3,48 @@
typedef struct a a;
struct a {
- a *b
+ a *b;
} d;
-e, k, ah, al;
-f(aa) {
+int e, k, ah, al;
+void h(void);
+void
+f(aa) int aa; {
if (aa & 1)
goto g;
f(aa | 2);
g:
h();
}
+void i();
+void
l() {
- {
f(072);
i(e, d, 92);
- }
}
+void
ag() {
- { i(e, d, 36); }
+ i(e, d, 36);
}
+void j();
+void
ai(a *m, a *n, unsigned aa) {
f(aa);
j(k, l, ah, 1);
}
+void
j(int c, a m, int aj, int aa) {
int ak = aa;
- { i(e, d, ak); }
+ i(e, d, ak);
}
+void
i(int c, a *m, unsigned aa) {
- {
- { i(c, (*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(
+ i(c, (*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(
*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(
*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*m).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b)
.b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b)
.b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b)
.b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b)
.b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b, 0);
- }
- }
int am = aa;
- ai(ag, al, am);
+ ai((a *) (void *) ag, (a *) (__INTPTR_TYPE__) al, am);
}
--- gcc/testsuite/g++.dg/ipa/pr93763.C.jj 2020-02-21 21:54:56.616348316 +0100
+++ gcc/testsuite/g++.dg/ipa/pr93763.C 2020-02-24 12:35:02.061484741 +0100
@@ -1,4 +1,4 @@
-/* { dg-do compile } */
+/* { dg-do compile { target c++11 } } */
/* { dg-options "-O3" } */
struct search_param {
Jakub
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] testsuite: Fix recently added ipa testcases [PR93763]
2020-02-24 11:50 ` [PATCH] testsuite: Fix recently added ipa testcases [PR93763] Jakub Jelinek
@ 2020-02-24 12:33 ` Christophe Lyon
0 siblings, 0 replies; 6+ messages in thread
From: Christophe Lyon @ 2020-02-24 12:33 UTC (permalink / raw)
To: Jakub Jelinek
Cc: Jeff Law, Martin Jambor, Feng Xue OS, Jan Hubicka, gcc-patches
On Mon, 24 Feb 2020 at 12:50, Jakub Jelinek <jakub@redhat.com> wrote:
>
> On Mon, Feb 24, 2020 at 11:09:58AM +0100, Christophe Lyon wrote:
> > I'm seeing errors with the new testcase:
> > end of struct or union
> > /gcc/testsuite/gcc.dg/ipa/pr93763.c:8:1: warning: data definition has
> > no type or storage class
> ...
>
> Seems the test has been badly reduced (if the original doesn't emit
> warnings, it is always better in the reduction script avoid introducing new
> ones).
> Also, the g++.dg/ipa/ test fails with -std=c++98 because it is written in
> C++11.
>
> Fixed thusly, ok for trunk?
Works for me, thanks
>
> 2020-02-24 Jakub Jelinek <jakub@redhat.com>
>
> PR ipa/93763
> * gcc.dg/ipa/pr93763.c: Adjust the test so that it compiles without
> warnings and still ICEs before the ipa-cp.c fix.
> * g++.dg/ipa/pr93763.C: Require c++11 effective target.
>
> --- gcc/testsuite/gcc.dg/ipa/pr93763.c.jj 2020-02-24 12:44:09.626326146 +0100
> +++ gcc/testsuite/gcc.dg/ipa/pr93763.c 2020-02-24 12:40:48.707320934 +0100
> @@ -3,44 +3,48 @@
>
> typedef struct a a;
> struct a {
> - a *b
> + a *b;
> } d;
> -e, k, ah, al;
> -f(aa) {
> +int e, k, ah, al;
> +void h(void);
> +void
> +f(aa) int aa; {
> if (aa & 1)
> goto g;
> f(aa | 2);
> g:
> h();
> }
> +void i();
> +void
> l() {
> - {
> f(072);
> i(e, d, 92);
> - }
> }
> +void
> ag() {
> - { i(e, d, 36); }
> + i(e, d, 36);
> }
> +void j();
> +void
> ai(a *m, a *n, unsigned aa) {
> f(aa);
> j(k, l, ah, 1);
> }
> +void
> j(int c, a m, int aj, int aa) {
> int ak = aa;
> - { i(e, d, ak); }
> + i(e, d, ak);
> }
> +void
> i(int c, a *m, unsigned aa) {
> - {
> - { i(c, (*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(
> + i(c, (*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(
> *(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(
> *(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*(*m).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b)
> .b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b)
> .b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b)
> .b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b)
> .b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b).b, 0);
> - }
> - }
> int am = aa;
> - ai(ag, al, am);
> + ai((a *) (void *) ag, (a *) (__INTPTR_TYPE__) al, am);
> }
> --- gcc/testsuite/g++.dg/ipa/pr93763.C.jj 2020-02-21 21:54:56.616348316 +0100
> +++ gcc/testsuite/g++.dg/ipa/pr93763.C 2020-02-24 12:35:02.061484741 +0100
> @@ -1,4 +1,4 @@
> -/* { dg-do compile } */
> +/* { dg-do compile { target c++11 } } */
> /* { dg-options "-O3" } */
>
> struct search_param {
>
>
> Jakub
>
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2020-02-24 12:33 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-02-18 15:12 [PATCH] Do not propagate self-dependent value (PR ipa/93763) Feng Xue OS
2020-02-21 17:59 ` Martin Jambor
2020-02-21 20:23 ` Jeff Law
2020-02-24 10:10 ` Christophe Lyon
2020-02-24 11:50 ` [PATCH] testsuite: Fix recently added ipa testcases [PR93763] Jakub Jelinek
2020-02-24 12:33 ` Christophe Lyon
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 read-only IMAP folder(s) and NNTP newsgroup(s).