[PATCH] gcc/config/tilegx/tilegx.c (tilegx_function_profiler): Save r10 to stack before call mcount

2016-05-29 Thread chengang 
From: Chen Gang 

r10 may also be as parameter for the nested function, so need save it
before call mcount.

2016-05-29  Chen Gang  

* config/tilegx/tilegx.c (tilegx_function_profiler): Save r10
to stack before call mcount.
---
 gcc/config/tilegx/tilegx.c | 14 --
 1 file changed, 12 insertions(+), 2 deletions(-)

diff --git a/gcc/config/tilegx/tilegx.c b/gcc/config/tilegx/tilegx.c
index 06c832c..bc41105 100644
--- a/gcc/config/tilegx/tilegx.c
+++ b/gcc/config/tilegx/tilegx.c
@@ -5510,18 +5510,28 @@ tilegx_function_profiler (FILE *file, int labelno 
ATTRIBUTE_UNUSED)
   if (flag_pic)
 {
   fprintf (file,
+  "\tst\tsp, r10\n"
+  "\taddi\tsp, sp, -8\n"
   "\t{\n"
   "\tmove\tr10, lr\n"
   "\tjal\tplt(%s)\n"
-  "\t}\n", MCOUNT_NAME);
+  "\t}\n"
+  "\taddi\tsp, sp, 8\n"
+  "\tld\tr10, sp\n",
+  MCOUNT_NAME);
 }
   else
 {
   fprintf (file,
+  "\tst\tsp, r10\n"
+  "\taddi\tsp, sp, -8\n"
   "\t{\n"
   "\tmove\tr10, lr\n"
   "\tjal\t%s\n"
-  "\t}\n", MCOUNT_NAME);
+  "\t}\n"
+  "\taddi\tsp, sp, 8\n"
+  "\tld\tr10, sp\n",
+  MCOUNT_NAME);
 }
 
   tilegx_in_bundle = false;
-- 
1.9.3

[PATCH v2] gcc/config/tilegx/tilegx.c (tilegx_function_profiler): Save r10 to stack before call mcount

2016-06-02 Thread chengang 
From: Chen Gang 

r10 may also be as parameter stack pointer for the nested function, so
need save it before call mcount.

2016-06-03  Chen Gang  

gcc/
PR target/71331
* config/tilegx/tilegx.c (tilegx_function_profiler): Save r10
to stack before call mcount.
---
 gcc/config/tilegx/tilegx.c | 14 --
 1 file changed, 12 insertions(+), 2 deletions(-)

diff --git a/gcc/config/tilegx/tilegx.c b/gcc/config/tilegx/tilegx.c
index 06c832c..bc41105 100644
--- a/gcc/config/tilegx/tilegx.c
+++ b/gcc/config/tilegx/tilegx.c
@@ -5510,18 +5510,32 @@ tilegx_function_profiler (FILE *file, int labelno 
ATTRIBUTE_UNUSED)
   if (flag_pic)
 {
   fprintf (file,
+  "\t{\n"
+  "\taddi\tsp, sp, -8\n"
+  "\tst\tsp, r10\n"
+  "\t}\n"
   "\t{\n"
   "\tmove\tr10, lr\n"
   "\tjal\tplt(%s)\n"
-  "\t}\n", MCOUNT_NAME);
+  "\t}\n"
+  "\taddi\tsp, sp, 8\n"
+  "\tld\tr10, sp\n",
+  MCOUNT_NAME);
 }
   else
 {
   fprintf (file,
+  "\t{\n"
+  "\taddi\tsp, sp, -8\n"
+  "\tst\tsp, r10\n"
+  "\t}\n"
   "\t{\n"
   "\tmove\tr10, lr\n"
   "\tjal\t%s\n"
-  "\t}\n", MCOUNT_NAME);
+  "\t}\n"
+  "\taddi\tsp, sp, 8\n"
+  "\tld\tr10, sp\n",
+  MCOUNT_NAME);
 }
 
   tilegx_in_bundle = false;
-- 
1.9.3

[PATCH v3] gcc/config/tilegx/tilegx.c (tilegx_function_profiler): Save r10 to stack before call mcount

2016-06-04 Thread chengang 
From: Chen Gang 

r10 may also be as parameter stack pointer for the nested function, so
need save it before call mcount.

Also clean up code: use '!' instead of "== 0" for checking
static_chain_decl and compute_total_frame_size.

2016-06-04  Chen Gang  

gcc/
PR target/71331
* config/tilegx/tilegx.c (tilegx_function_profiler): Save r10
to stack before call mcount.
(tilegx_can_use_return_insn_p): Clean up code.
---
 gcc/config/tilegx/tilegx.c | 20 ++--
 1 file changed, 18 insertions(+), 2 deletions(-)

diff --git a/gcc/config/tilegx/tilegx.c b/gcc/config/tilegx/tilegx.c
index 06c832c..55161bb 100644
--- a/gcc/config/tilegx/tilegx.c
+++ b/gcc/config/tilegx/tilegx.c
@@ -3880,8 +3880,8 @@ bool
 tilegx_can_use_return_insn_p (void)
 {
   return (reload_completed
- && cfun->static_chain_decl == 0
- && compute_total_frame_size () == 0
+ && !cfun->static_chain_decl
+ && !compute_total_frame_size ()
  && tilegx_current_function_is_leaf ()
  && !crtl->profile && !df_regs_ever_live_p (TILEGX_LINK_REGNUM));
 }
@@ -5507,6 +5507,15 @@ tilegx_function_profiler (FILE *file, int labelno 
ATTRIBUTE_UNUSED)
   fprintf (file, "\t}\n");
 }
 
+  if (cfun->static_chain_decl)
+{
+  fprintf (file,
+  "\t{\n"
+  "\taddi\tsp, sp, -8\n"
+  "\tst\tsp, r10\n"
+  "\t}\n");
+}
+
   if (flag_pic)
 {
   fprintf (file,
@@ -5524,6 +5533,13 @@ tilegx_function_profiler (FILE *file, int labelno 
ATTRIBUTE_UNUSED)
   "\t}\n", MCOUNT_NAME);
 }
 
+  if (cfun->static_chain_decl)
+{
+  fprintf (file,
+  "\taddi\tsp, sp, 8\n"
+  "\tld\tr10, sp\n");
+}
+
   tilegx_in_bundle = false;
 }
 
-- 
1.9.3

[PATCH] gcc: config: tilegx: Reserve prev insn when delete useless insn

2016-12-10 Thread chengang 
From: Chen Gang 

When check bundle, gcc may still need modify the prev insn. Original
implementation will lose the prev insn.

Also correct the coding styles of relate code.

2016-12-10  Chen Gang 

gcc/
PR target/78222
* tilegx.c (tilegx_gen_bundle): Reserve prev insn when delete
useless insn.
---
 gcc/config/tilegx/tilegx.c | 22 --
 1 file changed, 12 insertions(+), 10 deletions(-)

diff --git a/gcc/config/tilegx/tilegx.c b/gcc/config/tilegx/tilegx.c
index 76a7455..f846ca7 100644
--- a/gcc/config/tilegx/tilegx.c
+++ b/gcc/config/tilegx/tilegx.c
@@ -4469,8 +4469,7 @@ tilegx_gen_bundles (void)
   rtx_insn *end = NEXT_INSN (BB_END (bb));
 
   prev = NULL;
-  for (insn = next_insn_to_bundle (BB_HEAD (bb), end); insn;
-  prev = insn, insn = next)
+  for (insn = next_insn_to_bundle (BB_HEAD (bb), end); insn; insn = next)
{
  next = next_insn_to_bundle (NEXT_INSN (insn), end);
 
@@ -4499,14 +4498,17 @@ tilegx_gen_bundles (void)
  /* Delete barrier insns, because they can mess up the
 emitting of bundle braces.  If it is end-of-bundle, then
 the previous insn must be marked end-of-bundle.  */
- if (get_attr_type (insn) == TYPE_NOTHING) {
-   if (GET_MODE (insn) == QImode && prev != NULL
-   && GET_MODE (prev) == SImode)
- {
-   PUT_MODE (prev, QImode);
- }
-   delete_insn (insn);
- }
+ if (get_attr_type (insn) == TYPE_NOTHING)
+   {
+ if (GET_MODE (insn) == QImode && prev != NULL
+ && GET_MODE (prev) == SImode)
+   {
+ PUT_MODE (prev, QImode);
+   }
+   delete_insn (insn);
+   }
+ else
+   prev = insn;
}
 }
 }
-- 
2.5.0

Re: [PATCH] gcc: config: tilegx: Reserve prev insn when delete useless insn

2016-12-12 Thread chengang 
From: Chen Gang 

On Sat, Dec 10, 2016 at 03:38:10PM +, Bernd Edlinger wrote:
> On 12/10/16 12:23, cheng...@emindsoft.com.cn wrote:
> > From: Chen Gang 
> >
> > When check bundle, gcc may still need modify the prev insn. Original
> > implementation will lose the prev insn.
> >
> > Also correct the coding styles of relate code.
> >
> > 2016-12-10  Chen Gang 
> >
> > gcc/
> > PR target/78222
> > * tilegx.c (tilegx_gen_bundle): Reserve prev insn when delete
> > useless insn.
> 
> I think this was already fixed by Walt:
> 
> 
> r242617 | walt | 2016-11-19 03:34:17 +0100 (Sat, 19 Nov 2016) | 5 lines
> Changed paths:
> M /trunk/gcc/ChangeLog
> M /trunk/gcc/config/tilegx/tilegx.c
> 
> TILE-Gx: Fix bundling when encountering consecutive barriers.
> 
>  * config/tilegx/tilegx.c (tilegx_gen_bundles): Preserve
>end-of-bundle marker for consecutive barriers.
> 

Oh, yes, Good news to me.

> 
> But the formatting here is still odd, and should be fixed:
> TAB usage, single statements in braces, { not in a line of its own.
> 

Yes.

> I am however not sure about this statement:
> 
>/* Never wrap {} around inline asm.  */
>if (GET_CODE (PATTERN (insn)) != ASM_INPUT)
> 
> ... because this does only exclude asm(""); that is basic asm with
> empty assembler string.  To exclude all other forms of asm statements
> that are hidden in PARALLEL constructs we would need:
> 
>/* Never wrap {} around inline asm.  */
>if (GET_CODE (PATTERN (insn)) != ASM_INPUT
>&& asm_noperands (PATTERN (insn)) < 0)
> 
> 
> I think this if-condition is probably unnecessary, because it does
> apparently not create any problems although it is completely broken.
> 

For me, we need not touch it, since we are not quite sure about it. And
welcome any other related members' idea for it.


All together, I guess, I can leave this patch and continue to find and
send another patches about tilegx. If I should still do somthing about
this patch, please let me know.

Thanks.

>From cheng...@emindsoft.com.cn Mon Dec 12 19:23:58 2016
Date: Mon, 12 Dec 2016 19:23:58 +0800
From: Chen Gang 
To: Bernd Edlinger 
Cc: "cheng...@emindsoft.com.cn" ,
"l...@redhat.com" ,
"r...@redhat.com" ,
"mikest...@comcast.net" ,
"w...@tilera.com" ,
"gcc-patches@gcc.gnu.org" ,
"peter.mayd...@linaro.com" ,
"cmetc...@mellanox.com" 
Subject: Re: [PATCH] gcc: config: tilegx: Reserve prev insn when delete
 useless insn
Message-ID: <20161212112320.GA22787@localhost.localdomain>
References: <1481369032-26571-1-git-send-email-cheng...@emindsoft.com.cn>
 

MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: 

User-Agent: Mutt/1.7.1 (2016-10-04)
Status: RO
Content-Length: 2159
Lines: 66

On Sat, Dec 10, 2016 at 03:38:10PM +, Bernd Edlinger wrote:
> On 12/10/16 12:23, cheng...@emindsoft.com.cn wrote:
> > From: Chen Gang 
> >
> > When check bundle, gcc may still need modify the prev insn. Original
> > implementation will lose the prev insn.
> >
> > Also correct the coding styles of relate code.
> >
> > 2016-12-10  Chen Gang 
> >
> > gcc/
> > PR target/78222
> > * tilegx.c (tilegx_gen_bundle): Reserve prev insn when delete
> > useless insn.
> 
> I think this was already fixed by Walt:
> 
> 
> r242617 | walt | 2016-11-19 03:34:17 +0100 (Sat, 19 Nov 2016) | 5 lines
> Changed paths:
> M /trunk/gcc/ChangeLog
> M /trunk/gcc/config/tilegx/tilegx.c
> 
> TILE-Gx: Fix bundling when encountering consecutive barriers.
> 
>  * config/tilegx/tilegx.c (tilegx_gen_bundles): Preserve
>end-of-bundle marker for consecutive barriers.
> 

Oh, yes, Good news to me.

> 
> But the formatting here is still odd, and should be fixed:
> TAB usage, single statements in braces, { not in a line of its own.
> 

Yes.

> I am however not sure about this statement:
> 
>/* Never wrap {} around inline asm.  */
>if (GET_CODE (PATTERN (insn)) != ASM_INPUT)
> 
> ... because this does only exclude asm(""); that is basic asm with
> empty assembler string.  To exclude all other forms of asm statements
> that are hidden in PARALLEL constructs we would need:
> 
>/* Never wrap {} around inline asm.  */
>if (GET_CODE (PATTERN (insn)) != ASM_INPUT
>&& asm_noperands (PATTERN (insn)) < 0)
> 
> 
> I think this if-condition is probably unnecessary, because it does
> apparently not create any problems although it is completely broken.
> 

For me, we need not touch it, since we are not quite sure about it. And
welcome any other related members' idea for it.


All together, I guess, I can leave this patch and continue to find and
send another patches about tilegx. If I should still do somth

[PATCH v4] gcc/config/tilegx/tilegx.c (tilegx_function_profiler): Save r10 to stack before call mcount

2016-11-05 Thread chengang 
From: Chen Gang 

r10 may also be as parameter stack pointer for the nested function, so
need save it before call mcount.

Also clean up code: use '!' instead of "== 0" for checking
static_chain_decl and compute_total_frame_size.

2016-11-06  Chen Gang  

gcc/
PR target/71331
* config/tilegx/tilegx.c (tilegx_function_profiler): Save r10
to stack before call mcount.
(tilegx_can_use_return_insn_p): Clean up code.
---
 gcc/config/tilegx/tilegx.c | 20 ++--
 1 file changed, 18 insertions(+), 2 deletions(-)

diff --git a/gcc/config/tilegx/tilegx.c b/gcc/config/tilegx/tilegx.c
index e403c32..fd9181b 100644
--- a/gcc/config/tilegx/tilegx.c
+++ b/gcc/config/tilegx/tilegx.c
@@ -3919,8 +3919,8 @@ bool
 tilegx_can_use_return_insn_p (void)
 {
   return (reload_completed
- && cfun->static_chain_decl == 0
- && compute_total_frame_size () == 0
+ && !cfun->static_chain_decl
+ && !compute_total_frame_size ()
  && tilegx_current_function_is_leaf ()
  && !crtl->profile && !df_regs_ever_live_p (TILEGX_LINK_REGNUM));
 }
@@ -5551,6 +5551,15 @@ tilegx_function_profiler (FILE *file, int labelno 
ATTRIBUTE_UNUSED)
   fprintf (file, "\t}\n");
 }
 
+  if (cfun->static_chain_decl)
+{
+  fprintf (file,
+  "\t{\n"
+  "\taddi\tsp, sp, -16\n"
+  "\tst\tsp, r10\n"
+  "\t}\n");
+}
+
   if (flag_pic)
 {
   fprintf (file,
@@ -5568,6 +5577,13 @@ tilegx_function_profiler (FILE *file, int labelno 
ATTRIBUTE_UNUSED)
   "\t}\n", MCOUNT_NAME);
 }
 
+  if (cfun->static_chain_decl)
+{
+  fprintf (file,
+  "\taddi\tsp, sp, 16\n"
+  "\tld\tr10, sp\n");
+}
+
   tilegx_in_bundle = false;
 }
 
-- 
1.9.3