[Bug target/14798] New: In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[Bug target/14798] New: In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[anilp1 at kpitcummins dot com]

Hi,

I am using GCC 3.4 snapshot dated 17-03-2004.Please find following testcase 
which to produce the bug. 

test.c
*************************************************************
#pragma interrupt
void isr()
{
}
void delay(int a)
{
}
int main()
{
  return 0;
}
*************************************************************
Above test code is compiled with following options to gcc,
-m2e -S -O2

Since #pragma interrupt is written above "isr", so only "isr" should be treated 
as interrupt handler. But the functions mentioned below "isr" are also treated 
as interrupt handlers i.e. it saves all registers and generates "rte" instead 
of "rts". 

Compiler should treat only "isr" as interuupt handler.

Regards,
Anil Paranjpe

-- 
           Summary: In case of SH target with -O2 option #pragma interrupt
                    doesn't get resetted.
           Product: gcc
           Version: 3.4.0
            Status: UNCONFIRMED
          Severity: critical
          Priority: P1
         Component: target
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: anilp1 at kpitcummins dot com
                CC: anilp1 at kpitcummins dot com,gcc-bugs at gcc dot gnu
                    dot org
 GCC build triplet: i686-pc-linux-gnu
  GCC host triplet: i686-pc-linux-gnu
GCC target triplet: sh-unknown-elf


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/3.5 Regression] [unit-at-a-time?] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |wrong-code
            Summary|In case of SH target with - |[3.4/3.5 Regression] [unit-
                   |O2 option #pragma interrupt |at-a-time?] In case of SH
                   |doesn't get resetted.       |target with -O2 option
                   |                            |#pragma interrupt doesn't
                   |                            |get resetted.
   Target Milestone|---                         |3.4.1


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/3.5 Regression] [unit-at-a-time?] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[anilp1 at kpitcummins dot com]

------- Additional Comments From anilp1 at kpitcummins dot com  2004-04-15 12:46 -------
Subject: RE:  New: In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

Hi,

Please find following patch for bug no. 14798,

ChangeLog
2004-04-15	Anil Paranjpe <anilp1@kpitcummins.com>
	* sh/sh.c (sh_insert_attributes) : Flag pragma_interrupt is set to zero.

--- gcc/config/sh/sh.c.old	2004-04-15 17:25:30.000000000 +0530
+++ gcc/config/sh/sh.c	2004-04-15 17:36:08.000000000 +0530
@@ -6688,7 +6688,7 @@ sh_insert_attributes (tree node, tree *a
   if (! pragma_interrupt
       || TREE_CODE (node) != FUNCTION_DECL)
     return;
-
+    pragma_interrupt = 0;
   /* We are only interested in fields.  */
   if (TREE_CODE_CLASS (TREE_CODE (node)) != 'd')
     return;

Regards,
Anil Paranjpe

-----Original Message-----
From: anilp1 at kpitcummins dot com [mailto:gcc-bugzilla@gcc.gnu.org]
Sent: Wednesday, March 31, 2004 4:42 PM
To: Anil Paranjape
Subject: [Bug target/14798] New: In case of SH target with -O2 option
#pragma interrupt doesn't get resetted.


Hi,

I am using GCC 3.4 snapshot dated 17-03-2004.Please find following testcase 
which to produce the bug. 

test.c
*************************************************************
#pragma interrupt
void isr()
{
}
void delay(int a)
{
}
int main()
{
  return 0;
}
*************************************************************
Above test code is compiled with following options to gcc,
-m2e -S -O2

Since #pragma interrupt is written above "isr", so only "isr" should be treated 
as interrupt handler. But the functions mentioned below "isr" are also treated 
as interrupt handlers i.e. it saves all registers and generates "rte" instead 
of "rts". 

Compiler should treat only "isr" as interuupt handler.

Regards,
Anil Paranjpe

-- 
           Summary: In case of SH target with -O2 option #pragma interrupt
                    doesn't get resetted.
           Product: gcc
           Version: 3.4.0
            Status: UNCONFIRMED
          Severity: critical
          Priority: P1
         Component: target
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: anilp1 at kpitcummins dot com
                CC: anilp1 at kpitcummins dot com,gcc-bugs at gcc dot gnu
                    dot org
 GCC build triplet: i686-pc-linux-gnu
  GCC host triplet: i686-pc-linux-gnu
GCC target triplet: sh-unknown-elf


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

------- You are receiving this mail because: -------
You reported the bug, or are watching the reporter.
You are on the CC list for the bug, or are watching someone who is.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/3.5 Regression] [unit-at-a-time?] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-04-15 12:48 -------
Patch here: <http://gcc.gnu.org/ml/gcc-patches/2004-04/msg00893.html>. Confirmed.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
           Keywords|                            |patch
   Last reconfirmed|0000-00-00 00:00:00         |2004-04-15 12:48:40
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/3.5 Regression] [unit-at-a-time?] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[mmitchel at gcc dot gnu dot org]

------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-05-28 21:38 -------
Joern, Alexandre --

As Jim writes in his followup to Comment #2, there's messiness in the SH port
that is causing this regression.  As SH maintainers, would you two please take
care of this problem?

Thanks,

-- Mark

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |aoliva at redhat dot com,
                   |                            |joern dot rennecke at superh
                   |                            |dot com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/3.5 Regression] [unit-at-a-time?] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[aoliva at gcc dot gnu dot org]

------- Additional Comments From aoliva at gcc dot gnu dot org  2004-06-01 04:19 -------
Given the comment above the variable pragma_interrupt in sh.c, I assume the
behavior expected by Anil is the correct and intended behavior.  Although the
patch correctly implements this, I'd prefer to clear the variable only after
successfully adding the interrupt_handler attribute, right before returning. 
Patch approved with that change.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/3.5 Regression] [unit-at-a-time?] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[anilp1 at kpitcummins dot com]

------- Additional Comments From anilp1 at kpitcummins dot com  2004-06-01 06:49 -------
Subject: RE:  [3.4/3.5 Regression] [unit-at-a-time?] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

Hi all,

Please find patch updated with change suggested by Alexandre Oliva.

ChangeLog
2004-06-01	Anil Paranjpe <anilp1@kpitcummins.com>
	* sh/sh.c (sh_insert_attributes) : Reset pragma_interrupt after inserting interrupt_handler attribute.

Patch text
******************************************************************
--- gcc-3.4.0/gcc/config/sh/sh.c.orig	Sat May  1 16:40:57 2004
+++ gcc-3.4.0/gcc/config/sh/sh.c	Tue Jun  1 10:16:26 2004
@@ -6687,13 +6687,14 @@ sh_insert_attributes (tree node, tree *a
   if (! pragma_interrupt
       || TREE_CODE (node) != FUNCTION_DECL)
     return;
-
   /* We are only interested in fields.  */
   if (TREE_CODE_CLASS (TREE_CODE (node)) != 'd')
     return;
 
   /* Add a 'handle_interrupt' attribute.  */
   * attributes = tree_cons (get_identifier ("interrupt_handler"), NULL, * attributes);
+
+    pragma_interrupt = 0;
 
   return;
 }
******************************************************************

Please apply as i don't have access rights.

Regards,
Anil Paranjpe
KPIT Cummins InfoSystems Ltd.
Pune, India

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Free download of GNU based tool-chains for Renesas' SH and H8 Series.
The following site also offers free technical support to its users. 
Visit http://www.kpitgnutools.com for details. 
Latest versions of KPIT GNU tools are released on June 1, 2004.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 


-----Original Message-----
From: aoliva at gcc dot gnu dot org [mailto:gcc-bugzilla@gcc.gnu.org]
Sent: Tuesday, June 01, 2004 9:49 AM
To: Anil Paranjape
Subject: [Bug target/14798] [3.4/3.5 Regression] [unit-at-a-time?] In
case of SH target with -O2 option #pragma interrupt doesn't get
resetted.



------- Additional Comments From aoliva at gcc dot gnu dot org  2004-06-01 04:19 -------
Given the comment above the variable pragma_interrupt in sh.c, I assume the
behavior expected by Anil is the correct and intended behavior.  Although the
patch correctly implements this, I'd prefer to clear the variable only after
successfully adding the interrupt_handler attribute, right before returning. 
Patch approved with that change.



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/3.5 Regression] [unit-at-a-time?] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[wilson at specifixinc dot com]

------- Additional Comments From wilson at specifixinc dot com  2004-06-01 08:19 -------
Subject: Re:  [3.4/3.5 Regression] [unit-at-a-time?] In
 case of SH target with -O2 option #pragma interrupt doesn't get resetted.

aoliva at gcc dot gnu dot org wrote:
> ------- Additional Comments From aoliva at gcc dot gnu dot org  2004-06-01 04:19 -------
> Given the comment above the variable pragma_interrupt in sh.c, I assume the
> behavior expected by Anil is the correct and intended behavior.  Although the
> patch correctly implements this, I'd prefer to clear the variable only after
> successfully adding the interrupt_handler attribute, right before returning. 
> Patch approved with that change.

That just trades one set of problems for another set of problems.

Please see my review of the patch
     http://gcc.gnu.org/ml/gcc-patches/2004-04/msg01891.html


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/3.5 Regression] [unit-at-a-time?] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[aoliva at gcc dot gnu dot org]

------- Additional Comments From aoliva at gcc dot gnu dot org  2004-06-02 05:55 -------
Jim,

The points you bring up are good ones, and they do deserve bug reports.  But
this patch does fix the problem originally reported, and it is correct in itself.

Sure enough, other uses of this variable might be broken, but I don't see any
breakage.  Attribute handling is all done at parse time, and so is pragma
handling.  If this changes, we may have to rework this all, but it works as is
AFAICT.  Am I missing anything?

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/3.5 Regression] [unit-at-a-time?] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[wilson at specifixinc dot com]

------- Additional Comments From wilson at specifixinc dot com  2004-06-08 06:20 -------
Subject: Re:  [3.4/3.5 Regression] [unit-at-a-time?] In
 case of SH target with -O2 option #pragma interrupt doesn't get resetted.

aoliva at gcc dot gnu dot org wrote:
> The points you bring up are good ones, and they do deserve bug reports.  But
> this patch does fix the problem originally reported, and it is correct in itself.

I was arguing that the patch is wrong because while it fixes one 
attribute, it breaks another.  You are arguing that the patch is right 
because it fixes one attribute.  If you are happy with that, then that 
is fine with me.  It is your port.  I just feel an obligation to point 
out that there are still a number of other problems that haven't been 
fixed yet, and we risk losing knowledge of these other problems if we 
close this bug report.  But since this isn't my port, this isn't my problem.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/3.5 Regression] [unit-at-a-time?] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[anilp1 at kpitcummins dot com]

------- Additional Comments From anilp1 at kpitcummins dot com  2004-06-08 06:29 -------
Subject: RE:  [3.4/3.5 Regression] [unit-at-a-time?] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

Hi Jim,

Currently i have some other priorities, but i will certainly implement your suggestions as i find time.

Thanks and regards,
Anil

-----Original Message-----
From: wilson at specifixinc dot com [mailto:gcc-bugzilla@gcc.gnu.org]
Sent: Tuesday, June 08, 2004 11:51 AM
To: Anil Paranjape
Subject: [Bug target/14798] [3.4/3.5 Regression] [unit-at-a-time?] In
case of SH target with -O2 option #pragma interrupt doesn't get
resetted.



------- Additional Comments From wilson at specifixinc dot com  2004-06-08 06:20 -------
Subject: Re:  [3.4/3.5 Regression] [unit-at-a-time?] In
 case of SH target with -O2 option #pragma interrupt doesn't get resetted.

aoliva at gcc dot gnu dot org wrote:
> The points you bring up are good ones, and they do deserve bug reports.  But
> this patch does fix the problem originally reported, and it is correct in itself.

I was arguing that the patch is wrong because while it fixes one 
attribute, it breaks another.  You are arguing that the patch is right 
because it fixes one attribute.  If you are happy with that, then that 
is fine with me.  It is your port.  I just feel an obligation to point 
out that there are still a number of other problems that haven't been 
fixed yet, and we risk losing knowledge of these other problems if we 
close this bug report.  But since this isn't my port, this isn't my problem.




-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/3.5 Regression] [unit-at-a-time?] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[pinskia at gcc dot gnu dot org]

-- 
Bug 14798 depends on bug 15521, which changed state.

Bug 15521 Summary: [3.5 Regression] sched1 extends life of hard registers on SMALL_REGISTER_CLASSES targets when exceptions are enabled
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15521

           What    |Old Value                   |New Value
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/3.5 Regression] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[mmitchel at gcc dot gnu dot org]

------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-06-19 00:04 -------
Postponed until GCC 3.4.2.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|3.4.1                       |3.4.2


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/3.5 Regression] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[anilp1 at kpitcummins dot com]

------- Additional Comments From anilp1 at kpitcummins dot com  2004-07-12 10:17 -------
Subject: RE:  [3.4/3.5 Regression] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

Hi,
This problem is present in GCC 3.4.2 20040709 (prerelease) snapshot and also in GCC 3.5 20040704 snapshot. 

Can somebody tell me when it will be fixed ?

Regards,
Anil Paranjpe
KPIT Cummins InfoSystems Ltd.
Pune, India

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Free download of GNU based tool-chains for Renesas' SH and H8 Series.
The following site also offers free technical support to its users. 
Visit http://www.kpitgnutools.com for details. 
Latest versions of KPIT GNU tools were released on June 1, 2004.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 

-----Original Message-----
From: mmitchel at gcc dot gnu dot org [mailto:gcc-bugzilla@gcc.gnu.org]
Sent: Saturday, June 19, 2004 5:34 AM
To: Anil Paranjape
Subject: [Bug target/14798] [3.4/3.5 Regression] In case of SH target
with -O2 option #pragma interrupt doesn't get resetted.



------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-06-19 00:04 -------
Postponed until GCC 3.4.2.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|3.4.1                       |3.4.2


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

------- You are receiving this mail because: -------
You reported the bug, or are watching the reporter.
You are on the CC list for the bug, or are watching someone who is.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/3.5 Regression] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[steven at gcc dot gnu dot org]

------- Additional Comments From steven at gcc dot gnu dot org  2004-08-12 08:13 -------
The patch was shot down already. 

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|patch                       |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/3.5 Regression] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[mmitchel at gcc dot gnu dot org]

------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-08-23 20:56 -------
Postponed until GCC 3.4.3.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|3.4.2                       |3.4.3


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/4.0 Regression] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[mmitchel at gcc dot gnu dot org]

------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-10-31 02:05 -------
Postponed until GCC 3.4.4.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|3.4.3                       |3.4.4


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/4.0 Regression] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|critical                    |normal
           Priority|P1                          |P3


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/4.0 Regression] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-12-20 01:16 -------
Does the sh people care about this bug, it has been opened for almost 9 months now and no 
movement.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/4.0 Regression] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[mmitchel at gcc dot gnu dot org]

------- Additional Comments From mmitchel at gcc dot gnu dot org  2005-01-19 18:43 -------
SH is not a primary or secondary target; removing target milestone.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|3.4.4                       |---


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/4.0/4.1 Regression] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.1.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798

[Bug target/14798] [3.4/4.0/4.1 Regression] In case of SH target with -O2 option #pragma interrupt doesn't get resetted.

[mmitchel at gcc dot gnu dot org]

------- Additional Comments From mmitchel at gcc dot gnu dot org  2005-08-22 01:27 -------
SH is not a primary or secondary target; removing target milestone.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|4.1.0                       |---


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14798