Re: [PATCH][4.8 backport] Fix PR57735

Re: [PATCH][4.8 backport] Fix PR57735

[Richard Sandiford]

"Kyrylo Tkachov" <kyrylo.tkachov@arm.com> writes:
> Hi all,
>
> The fix for PR57735 is in current trunk (for a different issue I think), just
> needs a backport to 4.8.
> It is r198462 by Richard Sandiford:
>
> 2013-04-30 Richard Sandiford <rsandifo@linux.vnet.ibm.com>
>  	
>  	* explow.c (plus_constant): Pass "mode" to immed_double_int_const.
>  	Use gen_int_mode rather than GEN_INT.
>
> Ok to backport to the 4.8 branch?

Sorry for dropping the ball.  It's already been approved for 4.8,
and I think I even remembered to test it ready for commit.  I just
never actually hit commit afterwards :-(

Thanks,
Richard

Re: [PATCH][4.8 backport] Fix PR57735

[Richard Sandiford]

Richard Sandiford <rdsandiford@googlemail.com> writes:
> "Kyrylo Tkachov" <kyrylo.tkachov@arm.com> writes:
>> Hi all,
>>
>> The fix for PR57735 is in current trunk (for a different issue I think), just
>> needs a backport to 4.8.
>> It is r198462 by Richard Sandiford:
>>
>> 2013-04-30 Richard Sandiford <rsandifo@linux.vnet.ibm.com>
>>  	
>>  	* explow.c (plus_constant): Pass "mode" to immed_double_int_const.
>>  	Use gen_int_mode rather than GEN_INT.
>>
>> Ok to backport to the 4.8 branch?
>
> Sorry for dropping the ball.  It's already been approved for 4.8,
> and I think I even remembered to test it ready for commit.  I just
> never actually hit commit afterwards :-(

*sigh*.  Scratch that.  I'd confused it with:

2013-05-22  Richard Sandiford  <rsandifo@linux.vnet.ibm.com>

	* recog.c (offsettable_address_addr_space_p): Fix calculation of
	address mode.  Move pointer mode initialization to the same place.

which I _did_ apply to 4.8.  Sorry about the confusion instead...

Richard

RE: [PATCH][4.8 backport] Fix PR57735

[Kyrylo Tkachov]

Hi Richard,

> Richard Sandiford <rdsandiford@googlemail.com> writes:
> > "Kyrylo Tkachov" <kyrylo.tkachov@arm.com> writes:
> >> Hi all,
> >>
> >> The fix for PR57735 is in current trunk (for a different issue I
> think), just
> >> needs a backport to 4.8.
> >> It is r198462 by Richard Sandiford:
> >>
> >> 2013-04-30 Richard Sandiford <rsandifo@linux.vnet.ibm.com>
> >>
> >>  	* explow.c (plus_constant): Pass "mode" to immed_double_int_const.
> >>  	Use gen_int_mode rather than GEN_INT.
> >>
> >> Ok to backport to the 4.8 branch?
> >
> > Sorry for dropping the ball.  It's already been approved for 4.8,
> > and I think I even remembered to test it ready for commit.  I just
> > never actually hit commit afterwards :-(
> 
> *sigh*.  Scratch that.  I'd confused it with:
> 
> 2013-05-22  Richard Sandiford  <rsandifo@linux.vnet.ibm.com>
> 
> 	* recog.c (offsettable_address_addr_space_p): Fix calculation of
> 	address mode.  Move pointer mode initialization to the same place.
> 
> which I _did_ apply to 4.8.  Sorry about the confusion instead...

It's ok, didn't have time to act on the confusion anyway :)

So is the proposed one ok to backport? I've bootstrapped it on
arm-none-linux-gnueabihf and tested on arm-none-eabi with qemu.


Thanks,
Kyrill

RE: [PATCH][4.8 backport] Fix PR57735

[Kyrylo Tkachov]

Ping?

Thanks,
Kyrill

> -----Original Message-----
> From: gcc-patches-owner@gcc.gnu.org [mailto:gcc-patches-
> owner@gcc.gnu.org] On Behalf Of Kyrylo Tkachov
> Sent: 23 July 2013 10:09
> To: 'Richard Sandiford'
> Cc: gcc-patches; mikpe@it.uu.se; 'Richard Biener'
> Subject: RE: [PATCH][4.8 backport] Fix PR57735
> 
> Hi Richard,
> 
> > Richard Sandiford <rdsandiford@googlemail.com> writes:
> > > "Kyrylo Tkachov" <kyrylo.tkachov@arm.com> writes:
> > >> Hi all,
> > >>
> > >> The fix for PR57735 is in current trunk (for a different issue I
> > think), just
> > >> needs a backport to 4.8.
> > >> It is r198462 by Richard Sandiford:
> > >>
> > >> 2013-04-30 Richard Sandiford <rsandifo@linux.vnet.ibm.com>
> > >>
> > >>  	* explow.c (plus_constant): Pass "mode" to
> immed_double_int_const.
> > >>  	Use gen_int_mode rather than GEN_INT.
> > >>
> > >> Ok to backport to the 4.8 branch?
> > >
> > > Sorry for dropping the ball.  It's already been approved for 4.8,
> > > and I think I even remembered to test it ready for commit.  I just
> > > never actually hit commit afterwards :-(
> >
> > *sigh*.  Scratch that.  I'd confused it with:
> >
> > 2013-05-22  Richard Sandiford  <rsandifo@linux.vnet.ibm.com>
> >
> > 	* recog.c (offsettable_address_addr_space_p): Fix calculation of
> > 	address mode.  Move pointer mode initialization to the same place.
> >
> > which I _did_ apply to 4.8.  Sorry about the confusion instead...
> 
> It's ok, didn't have time to act on the confusion anyway :)
> 
> So is the proposed one ok to backport? I've bootstrapped it on
> arm-none-linux-gnueabihf and tested on arm-none-eabi with qemu.
> 
> 
> Thanks,
> Kyrill
> 
> 
> 
> 

RE: [PATCH][4.8 backport] Fix PR57735

[Kyrylo Tkachov]

> Ping?
> 

Ping^2

 Thanks,
 Kyrill
> 
> > -----Original Message-----
> > From: gcc-patches-owner@gcc.gnu.org [mailto:gcc-patches-
> > owner@gcc.gnu.org] On Behalf Of Kyrylo Tkachov
> > Sent: 23 July 2013 10:09
> > To: 'Richard Sandiford'
> > Cc: gcc-patches; mikpe@it.uu.se; 'Richard Biener'
> > Subject: RE: [PATCH][4.8 backport] Fix PR57735
> >
> > Hi Richard,
> >
> > > Richard Sandiford <rdsandiford@googlemail.com> writes:
> > > > "Kyrylo Tkachov" <kyrylo.tkachov@arm.com> writes:
> > > >> Hi all,
> > > >>
> > > >> The fix for PR57735 is in current trunk (for a different issue I
> > > think), just
> > > >> needs a backport to 4.8.
> > > >> It is r198462 by Richard Sandiford:
> > > >>
> > > >> 2013-04-30 Richard Sandiford <rsandifo@linux.vnet.ibm.com>
> > > >>
> > > >>  	* explow.c (plus_constant): Pass "mode" to
> > immed_double_int_const.
> > > >>  	Use gen_int_mode rather than GEN_INT.
> > > >>
> > > >> Ok to backport to the 4.8 branch?
> > > >
> > > > Sorry for dropping the ball.  It's already been approved for 4.8,
> > > > and I think I even remembered to test it ready for commit.  I just
> > > > never actually hit commit afterwards :-(
> > >
> > > *sigh*.  Scratch that.  I'd confused it with:
> > >
> > > 2013-05-22  Richard Sandiford  <rsandifo@linux.vnet.ibm.com>
> > >
> > > 	* recog.c (offsettable_address_addr_space_p): Fix calculation of
> > > 	address mode.  Move pointer mode initialization to the same place.
> > >
> > > which I _did_ apply to 4.8.  Sorry about the confusion instead...
> >
> > It's ok, didn't have time to act on the confusion anyway :)
> >
> > So is the proposed one ok to backport? I've bootstrapped it on
> > arm-none-linux-gnueabihf and tested on arm-none-eabi with qemu.
> >
> >
> > Thanks,
> > Kyrill
> >
> >
> >
> >
> 
> 
> 
> 

RE: [PATCH][4.8 backport] Fix PR57735

[Kyrylo Tkachov]

> > Ping?
> >
> 
> Ping^2
> 

Ping^3
http://gcc.gnu.org/ml/gcc-patches/2013-07/msg00929.html

>  Thanks,
>  Kyrill
> >
> > > -----Original Message-----
> > > From: gcc-patches-owner@gcc.gnu.org [mailto:gcc-patches-
> > > owner@gcc.gnu.org] On Behalf Of Kyrylo Tkachov
> > > Sent: 23 July 2013 10:09
> > > To: 'Richard Sandiford'
> > > Cc: gcc-patches; mikpe@it.uu.se; 'Richard Biener'
> > > Subject: RE: [PATCH][4.8 backport] Fix PR57735
> > >
> > > Hi Richard,
> > >
> > > > Richard Sandiford <rdsandiford@googlemail.com> writes:
> > > > > "Kyrylo Tkachov" <kyrylo.tkachov@arm.com> writes:
> > > > >> Hi all,
> > > > >>
> > > > >> The fix for PR57735 is in current trunk (for a different issue I
> > > > think), just
> > > > >> needs a backport to 4.8.
> > > > >> It is r198462 by Richard Sandiford:
> > > > >>
> > > > >> 2013-04-30 Richard Sandiford <rsandifo@linux.vnet.ibm.com>
> > > > >>
> > > > >>  	* explow.c (plus_constant): Pass "mode" to
> > > immed_double_int_const.
> > > > >>  	Use gen_int_mode rather than GEN_INT.
> > > > >>
> > > > >> Ok to backport to the 4.8 branch?
> > > > >
> > > > > Sorry for dropping the ball.  It's already been approved for 4.8,
> > > > > and I think I even remembered to test it ready for commit.  I
> just
> > > > > never actually hit commit afterwards :-(
> > > >
> > > > *sigh*.  Scratch that.  I'd confused it with:
> > > >
> > > > 2013-05-22  Richard Sandiford  <rsandifo@linux.vnet.ibm.com>
> > > >
> > > > 	* recog.c (offsettable_address_addr_space_p): Fix calculation
> of
> > > > 	address mode.  Move pointer mode initialization to the same
> place.
> > > >
> > > > which I _did_ apply to 4.8.  Sorry about the confusion instead...
> > >
> > > It's ok, didn't have time to act on the confusion anyway :)
> > >
> > > So is the proposed one ok to backport? I've bootstrapped it on
> > > arm-none-linux-gnueabihf and tested on arm-none-eabi with qemu.
> > >
> > >
> > > Thanks,
> > > Kyrill
> > >
> > >
> > >
> > >
> >
> >
> >
> >

Re: [PATCH][4.8 backport] Fix PR57735

[Richard Biener]

On Mon, Jul 22, 2013 at 6:27 PM, Kyrylo Tkachov <kyrylo.tkachov@arm.com> wrote:
> Hi all,
>
> The fix for PR57735 is in current trunk (for a different issue I think), just
> needs a backport to 4.8.
> It is r198462 by Richard Sandiford:
>
> 2013-04-30 Richard Sandiford <rsandifo@linux.vnet.ibm.com>
>
>         * explow.c (plus_constant): Pass "mode" to immed_double_int_const.
>         Use gen_int_mode rather than GEN_INT.
>
> Ok to backport to the 4.8 branch?
>
> I've attached the testcase that exposed the ICE. I think the ChangeLog would
> look like this:
>
>
> 2013-07-22  Kyrylo Tkachov  <kyrylo.tkachov@arm.com>
>
>         PR target/57735
>         Backport from mainline
>         2013-04-30  Richard Sandiford  <rsandifo@linux.vnet.ibm.com>
>
>                 * explow.c (plus_constant): Pass "mode" to
> immed_double_int_const.
>                 Use gen_int_mode rather than GEN_INT.
>
> 2013-07-22  Kyrylo Tkachov  <kyrylo.tkachov@arm.com>
>
>         PR target/57735
>         * g++.dg/ext/pr57735.C: New test.

Ok.  Please also add the testcase to the trunk.

Thanks,
Richard.

>
>
> Thanks,
> Kyrill

[PATCH][4.8 backport] Fix PR57735

[Kyrylo Tkachov]

[-- Attachment #1: Type: text/plain, Size: 879 bytes --]

Hi all,

The fix for PR57735 is in current trunk (for a different issue I think), just
needs a backport to 4.8.
It is r198462 by Richard Sandiford:

2013-04-30 Richard Sandiford <rsandifo@linux.vnet.ibm.com>
 	
 	* explow.c (plus_constant): Pass "mode" to immed_double_int_const.
 	Use gen_int_mode rather than GEN_INT.

Ok to backport to the 4.8 branch?

I've attached the testcase that exposed the ICE. I think the ChangeLog would
look like this:


2013-07-22  Kyrylo Tkachov  <kyrylo.tkachov@arm.com>

	PR target/57735
	Backport from mainline
	2013-04-30  Richard Sandiford  <rsandifo@linux.vnet.ibm.com>

		* explow.c (plus_constant): Pass "mode" to
immed_double_int_const.
		Use gen_int_mode rather than GEN_INT.
		
2013-07-22  Kyrylo Tkachov  <kyrylo.tkachov@arm.com>

	PR target/57735
	* g++.dg/ext/pr57735.C: New test.



Thanks,
Kyrill

[-- Attachment #2: pr57735-test.patch --]
[-- Type: application/octet-stream, Size: 4264 bytes --]

diff --git a/gcc/testsuite/g++.dg/ext/pr57735.C b/gcc/testsuite/g++.dg/ext/pr57735.C
new file mode 100644
index 0000000..0eb9500
--- /dev/null
+++ b/gcc/testsuite/g++.dg/ext/pr57735.C
@@ -0,0 +1,145 @@
+/* { dg-do compile { target arm*-*-* } } */
+/* { dg-options "-march=armv5te -marm  -mtune=xscale -mfloat-abi=soft -O1" } */
+
+typedef unsigned int size_t;
+__extension__
+typedef long long int int64_t;
+namespace WTF {
+    template<typename T> class RefPtr {
+    public:
+        inline T* operator->() const { return m_ptr; }
+        T* m_ptr;
+    };
+}
+using WTF::RefPtr;
+namespace JSC {
+    class ExecState;
+    class JSString;
+    typedef int64_t EncodedJSValue;
+    class JSValue {
+    public:
+        static EncodedJSValue encode(JSValue);
+        JSString* toString(ExecState*) const;
+    };
+}
+typedef unsigned char LChar;
+    typedef short unsigned int UChar;
+namespace WTF {
+    template<typename T, size_t inlineCapacity = 0>
+    class Vector {
+    public:
+        template<typename U> bool tryAppend(const U*, size_t);
+    };
+}
+using WTF::Vector;
+namespace WTF {
+template<typename CharType> inline bool isASCIIDigit(CharType c)
+{
+}
+template<typename CharType> inline bool isASCIIHexDigit(CharType c)
+{
+    return isASCIIDigit(c) || ((c | 0x20) >= 'a' && (c | 0x20) <= 'f');
+}
+    class StringImpl;
+}
+using WTF::StringImpl;
+namespace WTF {
+class StringImpl {
+public:
+    unsigned length() const { return m_length; }
+    unsigned m_length;
+};
+}
+namespace JSC {
+    class Register {
+    };
+class UString {
+public:
+    unsigned length() const
+    {
+        return m_impl->length();
+    }
+    const LChar* characters8() const
+    {
+    }
+    RefPtr<StringImpl> m_impl;
+};
+    class ExecState : private Register {
+    public:
+        JSValue argument(size_t argument)
+        {
+        }
+    };
+    class JSCell {
+    };
+    class JSString : public JSCell {
+    public:
+        const UString& value(ExecState*) const;
+    };
+class JSStringBuilder {
+public:
+    void append(const UChar u)
+    {
+        m_okay &= buffer16.tryAppend(&u, 1);
+    }
+    Vector<UChar, 64> buffer16;
+    bool m_okay;
+};
+template <typename T>
+class Lexer {
+public:
+    static unsigned char convertHex(int c1, int c2);
+};
+}
+namespace WTF {
+namespace Unicode {
+    int UTF8SequenceLength(char);
+    int decodeUTF8Sequence(const char*);
+}
+}
+using namespace WTF;
+using namespace Unicode;
+namespace JSC {
+template <typename CharType>
+static JSValue decode(ExecState* exec, const CharType* characters, int length, const char* doNotUnescape, bool strict)
+{
+    JSStringBuilder builder;
+    int k = 0;
+    UChar u = 0;
+    while (k < length) {
+        const CharType* p = characters + k;
+        CharType c = *p;
+        if (c == '%') {
+            int charLen = 0;
+            if (k <= length - 3 && isASCIIHexDigit(p[1]) && isASCIIHexDigit(p[2])) {
+                const char b0 = Lexer<CharType>::convertHex(p[1], p[2]);
+                const int sequenceLen = UTF8SequenceLength(b0);
+                if (sequenceLen && k <= length - sequenceLen * 3) {
+                    charLen = sequenceLen * 3;
+                    char sequence[5];
+                    if (charLen != 0) {
+                        const int character = decodeUTF8Sequence(sequence);
+                        if (character < 0 || character >= 0x110000)
+                            charLen = 0;
+                        else if (character >= 0x10000) {
+                            builder.append(static_cast<UChar>(0xD800 | ((character - 0x10000) >> 10)));
+                        } else
+                            u = static_cast<UChar>(character);
+                    }
+                }
+            }
+        }
+    }
+}
+static JSValue decode(ExecState* exec, const char* doNotUnescape, bool strict)
+{
+    UString str = exec->argument(0).toString(exec)->value(exec);
+        return decode(exec, str.characters8(), str.length(), doNotUnescape, strict);
+}
+EncodedJSValue globalFuncDecodeURI(ExecState* exec)
+{
+    static const char do_not_unescape_when_decoding_URI[] =
+        "#$&+,/:;=?@";
+    return JSValue::encode(decode(exec, do_not_unescape_when_decoding_URI, true));
+}
+}