diff mbox

[v2] gdb/i387-tdep.c: Avoid warning for "-Werror=strict-overflow"

Message ID 543A8208.9060605@gmail.com
State New
Headers show

Commit Message

Chen Gang Oct. 12, 2014, 1:28 p.m. UTC
gdb requires "-Werror", and I387_ST0_REGNUM (tdep) is 'variable', then
compiler can think that I387_ST0_REGNUM (tdep) may be a large number,
which may cause issue, so report warning.

Need fix this warning, and still keep the code clear enough for readers.
The related warning under Darwin with gnu built gcc:

  gcc -g -O2   -I. -I../../binutils-gdb/gdb -I../../binutils-gdb/gdb/common -I../../binutils-gdb/gdb/config -DLOCALEDIR="\"/usr/local/share/locale\"" -DHAVE_CONFIG_H -I../../binutils-gdb/gdb/../include/opcode -I../../binutils-gdb/gdb/../opcodes/..  -I../../binutils-gdb/gdb/../readline/.. -I../bfd -I../../binutils-gdb/gdb/../bfd -I../../binutils-gdb/gdb/../include -I../libdecnumber -I../../binutils-gdb/gdb/../libdecnumber -I../../binutils-gdb/gdb/gnulib/import -Ibuild-gnulib/import   -DTUI=1 -D_THREAD_SAFE -I/usr/local/Cellar/guile/2.0.11/include/guile/2.0 -I/usr/local/Cellar/gmp/6.0.0a/include -I/usr/local/Cellar/readline/6.3.5/include -I/usr/local/Cellar/bdw-gc/7.2e/include -I/System/Library/Frameworks/Python.framework/Versions/2.7/include/python2.7 -I/System/Library/Frameworks/Python.framework/Versions/2.7/include/python2.7 -Wall -Wdeclaration-after-statement -Wpointer-arith -Wpointer-sign -Wno-unused -Wunused-value -Wunused-function -Wno-switch -Wno-char-subscripts -Wmissing-protot
ypes -Wdeclaration-after-statement -Wempty-body -Wmissing-parameter-type -Wold-style-declaration -Wold-style-definition -Wformat-nonliteral -Werror -c -o i387-tdep.o -MT i387-tdep.o -MMD -MP -MF .deps/i387-tdep.Tpo ../../binutils-gdb/gdb/i387-tdep.c
  ../../binutils-gdb/gdb/i387-tdep.c: In function 'i387_supply_fsave':
  ../../binutils-gdb/gdb/i387-tdep.c:447:1: error: assuming signed overflow does not occur when assuming that (X + c) >= X is always true [-Werror=strict-overflow]
   i387_supply_fsave (struct regcache *regcache, int regnum, const void *fsave)
   ^
  ../../binutils-gdb/gdb/i387-tdep.c: In function 'i387_collect_fsave':
  ../../binutils-gdb/gdb/i387-tdep.c:502:1: error: assuming signed overflow does not occur when assuming that (X + c) >= X is always true [-Werror=strict-overflow]
   i387_collect_fsave (const struct regcache *regcache, int regnum, void *fsave)
   ^
  cc1: all warnings being treated as errors

2014-10-13  Chen Gang  <gang.chen.5i5j@gmail.com>

	* i387-tdep.c (i387_supply_fsave): Avoid warning for
	"-Werror=strict-overflow"
---
 gdb/i387-tdep.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

Comments

Iain Buclaw Oct. 12, 2014, 2:13 p.m. UTC | #1
On 12 October 2014 14:28, Chen Gang <gang.chen.5i5j@gmail.com> wrote:
> gdb requires "-Werror", and I387_ST0_REGNUM (tdep) is 'variable', then
> compiler can think that I387_ST0_REGNUM (tdep) may be a large number,
> which may cause issue, so report warning.
>
> Need fix this warning, and still keep the code clear enough for readers.
> The related warning under Darwin with gnu built gcc:
>

I had noted the same on GCC 5.0.0 development, found that the line
number in the warning was wrong and raised a bug
(https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63510), just didn't get
round to submitting a patch for gdb.


>
> 2014-10-13  Chen Gang  <gang.chen.5i5j@gmail.com>
>
>         * i387-tdep.c (i387_supply_fsave): Avoid warning for
>         "-Werror=strict-overflow"
> ---
>  gdb/i387-tdep.c | 10 ++++++----
>  1 file changed, 6 insertions(+), 4 deletions(-)
>
> diff --git a/gdb/i387-tdep.c b/gdb/i387-tdep.c
> index d66ac6a..609b5be 100644
> --- a/gdb/i387-tdep.c
> +++ b/gdb/i387-tdep.c
> @@ -450,11 +450,12 @@ i387_supply_fsave (struct regcache *regcache, int regnum, const void *fsave)
>    struct gdbarch_tdep *tdep = gdbarch_tdep (gdbarch);
>    enum bfd_endian byte_order = gdbarch_byte_order (gdbarch);
>    const gdb_byte *regs = fsave;
> -  int i;
> +  int i, end;
>

Also changing from int -> unsigned int fixes the warning.

-- Iain.
Chen Gang Oct. 12, 2014, 2:47 p.m. UTC | #2
On 10/12/14 22:13, Iain Buclaw wrote:
> On 12 October 2014 14:28, Chen Gang <gang.chen.5i5j@gmail.com> wrote:
>> gdb requires "-Werror", and I387_ST0_REGNUM (tdep) is 'variable', then
>> compiler can think that I387_ST0_REGNUM (tdep) may be a large number,
>> which may cause issue, so report warning.
>>
>> Need fix this warning, and still keep the code clear enough for readers.
>> The related warning under Darwin with gnu built gcc:
>>
> 
> I had noted the same on GCC 5.0.0 development, found that the line
> number in the warning was wrong and raised a bug
> (https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63510), just didn't get
> round to submitting a patch for gdb.
> 

But for me, what compiler has done is correct: "-Werror=strict-overflow"
need include "(X + c) >= X" for signed overflow. And our case matches
this case:

  int st0_regnum;
  
  void i387_supply_fsave ()
  {
    int i;
    for (i = st0_regnum; i < st0_regnum + 16;)
      {
      }
  }

> 
>>
>> 2014-10-13  Chen Gang  <gang.chen.5i5j@gmail.com>
>>
>>         * i387-tdep.c (i387_supply_fsave): Avoid warning for
>>         "-Werror=strict-overflow"
>> ---
>>  gdb/i387-tdep.c | 10 ++++++----
>>  1 file changed, 6 insertions(+), 4 deletions(-)
>>
>> diff --git a/gdb/i387-tdep.c b/gdb/i387-tdep.c
>> index d66ac6a..609b5be 100644
>> --- a/gdb/i387-tdep.c
>> +++ b/gdb/i387-tdep.c
>> @@ -450,11 +450,12 @@ i387_supply_fsave (struct regcache *regcache, int regnum, const void *fsave)
>>    struct gdbarch_tdep *tdep = gdbarch_tdep (gdbarch);
>>    enum bfd_endian byte_order = gdbarch_byte_order (gdbarch);
>>    const gdb_byte *regs = fsave;
>> -  int i;
>> +  int i, end;
>>
> 
> Also changing from int -> unsigned int fixes the warning.
> 
> -- Iain.
>
Iain Buclaw Oct. 12, 2014, 5:48 p.m. UTC | #3
On 12 October 2014 15:47, Chen Gang <gang.chen.5i5j@gmail.com> wrote:
> On 10/12/14 22:13, Iain Buclaw wrote:
>> On 12 October 2014 14:28, Chen Gang <gang.chen.5i5j@gmail.com> wrote:
>>> gdb requires "-Werror", and I387_ST0_REGNUM (tdep) is 'variable', then
>>> compiler can think that I387_ST0_REGNUM (tdep) may be a large number,
>>> which may cause issue, so report warning.
>>>
>>> Need fix this warning, and still keep the code clear enough for readers.
>>> The related warning under Darwin with gnu built gcc:
>>>
>>
>> I had noted the same on GCC 5.0.0 development, found that the line
>> number in the warning was wrong and raised a bug
>> (https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63510), just didn't get
>> round to submitting a patch for gdb.
>>
>
> But for me, what compiler has done is correct: "-Werror=strict-overflow"
> need include "(X + c) >= X" for signed overflow. And our case matches
> this case:
>

The compiler has done right, but that still doesn't stop the reported
line number being wrong.

-- Iain.
Mark Kettenis Oct. 13, 2014, 3:16 p.m. UTC | #4
> Date: Sun, 12 Oct 2014 21:28:40 +0800
> From: Chen Gang <gang.chen.5i5j@gmail.com>
> 
> gdb requires "-Werror", and I387_ST0_REGNUM (tdep) is 'variable', then
> compiler can think that I387_ST0_REGNUM (tdep) may be a large number,
> which may cause issue, so report warning.
> 
> Need fix this warning, and still keep the code clear enough for readers.
> The related warning under Darwin with gnu built gcc:

Sorry, I disagree with this approach.  I think your compiler is really
being unhelpful here.  See if a newer GCC version of the compiler
still triggers that warning.  If so, complain to the GCC people.

>   gcc -g -O2   -I. -I../../binutils-gdb/gdb -I../../binutils-gdb/gdb/common -I../../binutils-gdb/gdb/config -DLOCALEDIR="\"/usr/local/share/locale\"" -DHAVE_CONFIG_H -I../../binutils-gdb/gdb/../include/opcode -I../../binutils-gdb/gdb/../opcodes/..  -I../../binutils-gdb/gdb/../readline/.. -I../bfd -I../../binutils-gdb/gdb/../bfd -I../../binutils-gdb/gdb/../include -I../libdecnumber -I../../binutils-gdb/gdb/../libdecnumber -I../../binutils-gdb/gdb/gnulib/import -Ibuild-gnulib/import   -DTUI=1 -D_THREAD_SAFE -I/usr/local/Cellar/guile/2.0.11/include/guile/2.0 -I/usr/local/Cellar/gmp/6.0.0a/include -I/usr/local/Cellar/readline/6.3.5/include -I/usr/local/Cellar/bdw-gc/7.2e/include -I/System/Library/Frameworks/Python.framework/Versions/2.7/include/python2.7 -I/System/Library/Frameworks/Python.framework/Versions/2.7/include/python2.7 -Wall -Wdeclaration-after-statement -Wpointer-arith -Wpointer-sign -Wno-unused -Wunused-value -Wunused-function -Wno-switch -Wno-char-subscripts -Wmi!
 ssing-protot
> ypes -Wdeclaration-after-statement -Wempty-body -Wmissing-parameter-type -Wold-style-declaration -Wold-style-definition -Wformat-nonliteral -Werror -c -o i387-tdep.o -MT i387-tdep.o -MMD -MP -MF .deps/i387-tdep.Tpo ../../binutils-gdb/gdb/i387-tdep.c
>   ../../binutils-gdb/gdb/i387-tdep.c: In function 'i387_supply_fsave':
>   ../../binutils-gdb/gdb/i387-tdep.c:447:1: error: assuming signed overflow does not occur when assuming that (X + c) >= X is always true [-Werror=strict-overflow]
>    i387_supply_fsave (struct regcache *regcache, int regnum, const void *fsave)
>    ^
>   ../../binutils-gdb/gdb/i387-tdep.c: In function 'i387_collect_fsave':
>   ../../binutils-gdb/gdb/i387-tdep.c:502:1: error: assuming signed overflow does not occur when assuming that (X + c) >= X is always true [-Werror=strict-overflow]
>    i387_collect_fsave (const struct regcache *regcache, int regnum, void *fsave)
>    ^
>   cc1: all warnings being treated as errors
> 
> 2014-10-13  Chen Gang  <gang.chen.5i5j@gmail.com>
> 
> 	* i387-tdep.c (i387_supply_fsave): Avoid warning for
> 	"-Werror=strict-overflow"
> ---
>  gdb/i387-tdep.c | 10 ++++++----
>  1 file changed, 6 insertions(+), 4 deletions(-)
> 
> diff --git a/gdb/i387-tdep.c b/gdb/i387-tdep.c
> index d66ac6a..609b5be 100644
> --- a/gdb/i387-tdep.c
> +++ b/gdb/i387-tdep.c
> @@ -450,11 +450,12 @@ i387_supply_fsave (struct regcache *regcache, int regnum, const void *fsave)
>    struct gdbarch_tdep *tdep = gdbarch_tdep (gdbarch);
>    enum bfd_endian byte_order = gdbarch_byte_order (gdbarch);
>    const gdb_byte *regs = fsave;
> -  int i;
> +  int i, end;
>  
>    gdb_assert (tdep->st0_regnum >= I386_ST0_REGNUM);
>  
> -  for (i = I387_ST0_REGNUM (tdep); i < I387_XMM0_REGNUM (tdep); i++)
> +  end = I387_XMM0_REGNUM (tdep); /* -Werror=strict-overflow for (X + c) >= X */
> +  for (i = I387_ST0_REGNUM (tdep); i < end; i++)
>      if (regnum == -1 || regnum == i)
>        {
>  	if (fsave == NULL)
> @@ -503,11 +504,12 @@ i387_collect_fsave (const struct regcache *regcache, int regnum, void *fsave)
>  {
>    struct gdbarch_tdep *tdep = gdbarch_tdep (get_regcache_arch (regcache));
>    gdb_byte *regs = fsave;
> -  int i;
> +  int i, end;
>  
>    gdb_assert (tdep->st0_regnum >= I386_ST0_REGNUM);
>  
> -  for (i = I387_ST0_REGNUM (tdep); i < I387_XMM0_REGNUM (tdep); i++)
> +  end = I387_XMM0_REGNUM (tdep); /* -Werror=strict-overflow for (X + c) >= X */
> +  for (i = I387_ST0_REGNUM (tdep); i < end; i++)
>      if (regnum == -1 || regnum == i)
>        {
>  	/* Most of the FPU control registers occupy only 16 bits in
> -- 
> 1.8.5.2 (Apple Git-48)
>
Chen Gang Oct. 13, 2014, 4:05 p.m. UTC | #5
On 10/13/14 23:16, Mark Kettenis wrote:
>> Date: Sun, 12 Oct 2014 21:28:40 +0800
>> From: Chen Gang <gang.chen.5i5j@gmail.com>
>>
>> gdb requires "-Werror", and I387_ST0_REGNUM (tdep) is 'variable', then
>> compiler can think that I387_ST0_REGNUM (tdep) may be a large number,
>> which may cause issue, so report warning.
>>
>> Need fix this warning, and still keep the code clear enough for readers.
>> The related warning under Darwin with gnu built gcc:
> 
> Sorry, I disagree with this approach.  I think your compiler is really
> being unhelpful here.  See if a newer GCC version of the compiler
> still triggers that warning.  If so, complain to the GCC people.
> 

Excuse me, I do not dare to complain it to gcc members, because for me,
the compiler's printing is obviously, and proofs itself is correct. If I
ask a quite obvious question to them, I guess, I will be complained.

Hmm... But I may have an additional trying: Use original gcc and provide
"-Werror=strict-overflow" to see what will happen.

 - If it will report warning too, I can ask gcc members why integrate
   "-Wstrict-overflow" to -Wall.

 - If it will not report warning, it means original gcc also integrate
   "-Wstrict-overflow" or have no this option. And for me, I prefer to
   use "#pragma GCC diagnostic ignored ... " for our code.

Thanks.

>>   gcc -g -O2   -I. -I../../binutils-gdb/gdb -I../../binutils-gdb/gdb/common -I../../binutils-gdb/gdb/config -DLOCALEDIR="\"/usr/local/share/locale\"" -DHAVE_CONFIG_H -I../../binutils-gdb/gdb/../include/opcode -I../../binutils-gdb/gdb/../opcodes/..  -I../../binutils-gdb/gdb/../readline/.. -I../bfd -I../../binutils-gdb/gdb/../bfd -I../../binutils-gdb/gdb/../include -I../libdecnumber -I../../binutils-gdb/gdb/../libdecnumber -I../../binutils-gdb/gdb/gnulib/import -Ibuild-gnulib/import   -DTUI=1 -D_THREAD_SAFE -I/usr/local/Cellar/guile/2.0.11/include/guile/2.0 -I/usr/local/Cellar/gmp/6.0.0a/include -I/usr/local/Cellar/readline/6.3.5/include -I/usr/local/Cellar/bdw-gc/7.2e/include -I/System/Library/Frameworks/Python.framework/Versions/2.7/include/python2.7 -I/System/Library/Frameworks/Python.framework/Versions/2.7/include/python2.7 -Wall -Wdeclaration-after-statement -Wpointer-arith -Wpointer-sign -Wno-unused -Wunused-value -Wunused-function -Wno-switch -Wno-char-subscripts -Wmi!
>  ssing-protot
>> ypes -Wdeclaration-after-statement -Wempty-body -Wmissing-parameter-type -Wold-style-declaration -Wold-style-definition -Wformat-nonliteral -Werror -c -o i387-tdep.o -MT i387-tdep.o -MMD -MP -MF .deps/i387-tdep.Tpo ../../binutils-gdb/gdb/i387-tdep.c
>>   ../../binutils-gdb/gdb/i387-tdep.c: In function 'i387_supply_fsave':
>>   ../../binutils-gdb/gdb/i387-tdep.c:447:1: error: assuming signed overflow does not occur when assuming that (X + c) >= X is always true [-Werror=strict-overflow]
>>    i387_supply_fsave (struct regcache *regcache, int regnum, const void *fsave)
>>    ^
>>   ../../binutils-gdb/gdb/i387-tdep.c: In function 'i387_collect_fsave':
>>   ../../binutils-gdb/gdb/i387-tdep.c:502:1: error: assuming signed overflow does not occur when assuming that (X + c) >= X is always true [-Werror=strict-overflow]
>>    i387_collect_fsave (const struct regcache *regcache, int regnum, void *fsave)
>>    ^
>>   cc1: all warnings being treated as errors
>>
>> 2014-10-13  Chen Gang  <gang.chen.5i5j@gmail.com>
>>
>> 	* i387-tdep.c (i387_supply_fsave): Avoid warning for
>> 	"-Werror=strict-overflow"
>> ---
>>  gdb/i387-tdep.c | 10 ++++++----
>>  1 file changed, 6 insertions(+), 4 deletions(-)
>>
>> diff --git a/gdb/i387-tdep.c b/gdb/i387-tdep.c
>> index d66ac6a..609b5be 100644
>> --- a/gdb/i387-tdep.c
>> +++ b/gdb/i387-tdep.c
>> @@ -450,11 +450,12 @@ i387_supply_fsave (struct regcache *regcache, int regnum, const void *fsave)
>>    struct gdbarch_tdep *tdep = gdbarch_tdep (gdbarch);
>>    enum bfd_endian byte_order = gdbarch_byte_order (gdbarch);
>>    const gdb_byte *regs = fsave;
>> -  int i;
>> +  int i, end;
>>  
>>    gdb_assert (tdep->st0_regnum >= I386_ST0_REGNUM);
>>  
>> -  for (i = I387_ST0_REGNUM (tdep); i < I387_XMM0_REGNUM (tdep); i++)
>> +  end = I387_XMM0_REGNUM (tdep); /* -Werror=strict-overflow for (X + c) >= X */
>> +  for (i = I387_ST0_REGNUM (tdep); i < end; i++)
>>      if (regnum == -1 || regnum == i)
>>        {
>>  	if (fsave == NULL)
>> @@ -503,11 +504,12 @@ i387_collect_fsave (const struct regcache *regcache, int regnum, void *fsave)
>>  {
>>    struct gdbarch_tdep *tdep = gdbarch_tdep (get_regcache_arch (regcache));
>>    gdb_byte *regs = fsave;
>> -  int i;
>> +  int i, end;
>>  
>>    gdb_assert (tdep->st0_regnum >= I386_ST0_REGNUM);
>>  
>> -  for (i = I387_ST0_REGNUM (tdep); i < I387_XMM0_REGNUM (tdep); i++)
>> +  end = I387_XMM0_REGNUM (tdep); /* -Werror=strict-overflow for (X + c) >= X */
>> +  for (i = I387_ST0_REGNUM (tdep); i < end; i++)
>>      if (regnum == -1 || regnum == i)
>>        {
>>  	/* Most of the FPU control registers occupy only 16 bits in
>> -- 
>> 1.8.5.2 (Apple Git-48)
>>
Chen Gang Oct. 14, 2014, 9:30 p.m. UTC | #6
On 10/14/2014 12:05 AM, Chen Gang wrote:
> On 10/13/14 23:16, Mark Kettenis wrote:
>>> Date: Sun, 12 Oct 2014 21:28:40 +0800
>>> From: Chen Gang <gang.chen.5i5j@gmail.com>
>>>
>>> gdb requires "-Werror", and I387_ST0_REGNUM (tdep) is 'variable', then
>>> compiler can think that I387_ST0_REGNUM (tdep) may be a large number,
>>> which may cause issue, so report warning.
>>>
>>> Need fix this warning, and still keep the code clear enough for readers.
>>> The related warning under Darwin with gnu built gcc:
>>
>> Sorry, I disagree with this approach.  I think your compiler is really
>> being unhelpful here.  See if a newer GCC version of the compiler
>> still triggers that warning.  If so, complain to the GCC people.
>>
> 
> Excuse me, I do not dare to complain it to gcc members, because for me,
> the compiler's printing is obviously, and proofs itself is correct. If I
> ask a quite obvious question to them, I guess, I will be complained.
> 
> Hmm... But I may have an additional trying: Use original gcc and provide
> "-Werror=strict-overflow" to see what will happen.
> 
>  - If it will report warning too, I can ask gcc members why integrate
>    "-Wstrict-overflow" to -Wall.
> 
>  - If it will not report warning, it means original gcc also integrate
>    "-Wstrict-overflow" or have no this option. And for me, I prefer to
>    use "#pragma GCC diagnostic ignored ... " for our code.

After use "gcc version 4.8.3 20140624 (Red Hat 4.8.3-1) (GCC)" with
"-Wstrict-overflow", it supports this flag, and does not report warning.

For me, either "$pragma GCC diagnostic ignored ..." or continuing this
patch. At present, I don't know any other ways for it.

Thanks.
Joel Brobecker Oct. 15, 2014, 6:47 p.m. UTC | #7
Hi Mark,

> > Sorry, I disagree with this approach.  I think your compiler is really
> > being unhelpful here.  See if a newer GCC version of the compiler
> > still triggers that warning.  If so, complain to the GCC people.

Would you be OK with Pedro's suggested approach where we have
an additional macro that provides the number of ST registers?
That should get rid of the warning, and it seems like the code
might actually even be slightly better for it; it would make it
obvious that you iterate over a fixed number of registers, and
which ones you are interating over; [ST0 .. XMM0[ does not make
it as clear that the range only refers to ST registers, for instance.
Mark Kettenis Oct. 15, 2014, 6:58 p.m. UTC | #8
> Date: Wed, 15 Oct 2014 11:47:11 -0700
> From: Joel Brobecker <brobecker@adacore.com>
> 
> Hi Mark,
> 
> > > Sorry, I disagree with this approach.  I think your compiler is really
> > > being unhelpful here.  See if a newer GCC version of the compiler
> > > still triggers that warning.  If so, complain to the GCC people.
> 
> Would you be OK with Pedro's suggested approach where we have
> an additional macro that provides the number of ST registers?
> That should get rid of the warning, and it seems like the code
> might actually even be slightly better for it; it would make it
> obvious that you iterate over a fixed number of registers, and
> which ones you are interating over; [ST0 .. XMM0[ does not make
> it as clear that the range only refers to ST registers, for instance.

Pedro's suggestion made sense to me.
diff mbox

Patch

diff --git a/gdb/i387-tdep.c b/gdb/i387-tdep.c
index d66ac6a..609b5be 100644
--- a/gdb/i387-tdep.c
+++ b/gdb/i387-tdep.c
@@ -450,11 +450,12 @@  i387_supply_fsave (struct regcache *regcache, int regnum, const void *fsave)
   struct gdbarch_tdep *tdep = gdbarch_tdep (gdbarch);
   enum bfd_endian byte_order = gdbarch_byte_order (gdbarch);
   const gdb_byte *regs = fsave;
-  int i;
+  int i, end;
 
   gdb_assert (tdep->st0_regnum >= I386_ST0_REGNUM);
 
-  for (i = I387_ST0_REGNUM (tdep); i < I387_XMM0_REGNUM (tdep); i++)
+  end = I387_XMM0_REGNUM (tdep); /* -Werror=strict-overflow for (X + c) >= X */
+  for (i = I387_ST0_REGNUM (tdep); i < end; i++)
     if (regnum == -1 || regnum == i)
       {
 	if (fsave == NULL)
@@ -503,11 +504,12 @@  i387_collect_fsave (const struct regcache *regcache, int regnum, void *fsave)
 {
   struct gdbarch_tdep *tdep = gdbarch_tdep (get_regcache_arch (regcache));
   gdb_byte *regs = fsave;
-  int i;
+  int i, end;
 
   gdb_assert (tdep->st0_regnum >= I386_ST0_REGNUM);
 
-  for (i = I387_ST0_REGNUM (tdep); i < I387_XMM0_REGNUM (tdep); i++)
+  end = I387_XMM0_REGNUM (tdep); /* -Werror=strict-overflow for (X + c) >= X */
+  for (i = I387_ST0_REGNUM (tdep); i < end; i++)
     if (regnum == -1 || regnum == i)
       {
 	/* Most of the FPU control registers occupy only 16 bits in