error processing 220-noteGNUstack-01.patch

error processing 220-noteGNUstack-01.patch

[ng]

HI,

I just got the new 1.4.0 release and tried to set up with gcc-4.3.2

[DEBUG]    ==> Executing: 'patch -g0 -F1 -p1 -f'
[ALL  ]    The next patch would create the file
boehm-gc/ia64_save_regs_in_stack.S,
[ALL  ]    which already exists!  Applying it anyway.
[ALL  ]    patching file boehm-gc/ia64_save_regs_in_stack.S
[ALL  ]    Patch attempted to create file
boehm-gc/ia64_save_regs_in_stack.S, which already exists.
[ALL  ]    Hunk #1 FAILED at 1.
[ALL  ]    1 out of 1 hunk FAILED -- saving rejects to file
boehm-gc/ia64_save_regs_in_stack.S.rej
[ALL  ]    patching file boehm-gc/ia64_save_regs_in_stack.s
[ALL  ]    patching file libffi/src/alpha/osf.S



since I'm building for ARM on x86 I  suppose I could just edit the patch 
but is this a known problem or a slip up somewhere?

I'm still finding my feet with this tool but it looks like an excellent 
and flexible way to build a cross toolchain that takes a lot of the pain 
out of the process.

Thanks.

--
For unsubscribe information see http://sourceware.org/lists.html#faq

Re: error processing 220-noteGNUstack-01.patch

[Yann E. MORIN]

Hello!

On Wednesday 29 April 2009 21:08:51 ng@piments.com wrote:
> I just got the new 1.4.0 release and tried to set up with gcc-4.3.2
> [DEBUG]    ==> Executing: 'patch -g0 -F1 -p1 -f'
> [ALL  ]    The next patch would create the file
> boehm-gc/ia64_save_regs_in_stack.S,
> [ALL  ]    which already exists!  Applying it anyway.

Oh no! Not again! ;-(

I thought it had been sorted out before the release...

> since I'm building for ARM on x86 I  suppose I could just edit the patch 
> but is this a known problem or a slip up somewhere?

It already happened in the past, and the patch was updated.
Looks like something went wrong... Let me browse the code...
Nope the patch is fine.

Are you building on Linux or Windows?
What is the file system you're building on?

Regards,
Yann E. MORIN.

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +0/33 662376056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| --==< ^_^ >==-- `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
`------------------------------^-------^------------------^--------------------'


--
For unsubscribe information see http://sourceware.org/lists.html#faq

Re: error processing 220-noteGNUstack-01.patch

[Yann E. MORIN]

Hello!

Please, keep the mailing list copied when you answer.

On Wednesday 29 April 2009 21:42:32 ng@piments.com wrote:
> Yann E. MORIN wrote:
> > Are you building on Linux or Windows?
> > What is the file system you're building on?
> gentoo linux, reiser4

OK. I guess reiser4 is case-sensitive, is it not?

The only reason (that I know of) which would make the patch fail is
a case-insensitive file system (such as Windows FAT32).

Regards,
Yann E. MORIN.

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +0/33 662376056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| --==< ^_^ >==-- `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
`------------------------------^-------^------------------^--------------------'


--
For unsubscribe information see http://sourceware.org/lists.html#faq

Re: error processing 220-noteGNUstack-01.patch

[ng]

Yann E. MORIN wrote:
> Hello!
> 
> Please, keep the mailing list copied when you answer.
> 
> On Wednesday 29 April 2009 21:42:32 ng@piments.com wrote:
>> Yann E. MORIN wrote:
>>> Are you building on Linux or Windows?
>>> What is the file system you're building on?
>> gentoo linux, reiser4
> 
> OK. I guess reiser4 is case-sensitive, is it not?
> 
> The only reason (that I know of) which would make the patch fail is
> a case-insensitive file system (such as Windows FAT32).
> 
> Regards,
> Yann E. MORIN.
> 

Sure, it's case sensitive.

but this does not seem to be a false error , the file does already 
exist. I have not rerun anything since the run that failed. The error it 
states is correct. The file exists and has the expected content. Why is 
the patch being called?

$ ls  boehm-gc/ia64_save_regs_in_stack.S
boehm-gc/ia64_save_regs_in_stack.S

$ cat !$
cat boehm-gc/ia64_save_regs_in_stack.S
         .text
         .align 16
         .global GC_save_regs_in_stack
         .proc GC_save_regs_in_stack
GC_save_regs_in_stack:
         .bodyfoo.mpg
         flushrs
         ;;
         mov r8=ar.bsp
         br.ret.sptk.few rp
         .endp GC_save_regs_in_stack

#ifdef __linux__
         .section .note.GNU-stack,"",@progbits
#endif


Is there a way with ct-ng to remove this patch and rebuild without it 
being refeched?

Or can you suggest another work around?

Thanks for your help.

--
For unsubscribe information see http://sourceware.org/lists.html#faq

Re: error processing 220-noteGNUstack-01.patch

[ng]

Yann E. MORIN wrote:
> Hello!
> 
> Please, keep the mailing list copied when you answer.
> 
> On Wednesday 29 April 2009 21:42:32 ng@piments.com wrote:
>> Yann E. MORIN wrote:
>>> Are you building on Linux or Windows?
>>> What is the file system you're building on?
>> gentoo linux, reiser4
> 
> OK. I guess reiser4 is case-sensitive, is it not?
> 
> The only reason (that I know of) which would make the patch fail is
> a case-insensitive file system (such as Windows FAT32).
> 
> Regards,
> Yann E. MORIN.
> 

Well I did the clean up as noted in the output and removed the 200* 
patch and checking the file that would be created by this patch is there 
, so it is in the tarball already.

This got me passed the failed hunk but I hit a similar problem in gdb , 
again the file exists.

Is this not simply that some of these patches should not be in the more 
recent versions patchsets?



[ALL  ]    patching file gdb/testsuite/gdb.base/gdbinit.exp
[ALL  ]    Patch attempted to create file 
gdb/testsuite/gdb.base/gdbinit.exp, which already exists.
[ALL  ]    Hunk #1 FAILED at 1.
[ALL  ]    1 out of 1 hunk FAILED -- saving rejects to file 
gdb/testsuite/gdb.base/gdbinit.exp.rej
[ALL  ]    The next patch would create the file 
gdb/testsuite/gdb.base/gdbinit.sample,
[ALL  ]    which already exists!  Applying it anyway.
[ALL  ]    patching file gdb/testsuite/gdb.base/gdbinit.sample
[ALL  ]    Patch attempted to create file 
gdb/testsuite/gdb.base/gdbinit.sample, which already exists.
[ALL  ]    Hunk #1 FAILED at 1.
[ALL  ]    1 out of 1 hunk FAILED -- saving rejects to file 
gdb/testsuite/gdb.base/gdbinit.sample.rej
[ERROR]    Build failed in step 'Extracting and patching toolchain 
components'

Thanks again.


--
For unsubscribe information see http://sourceware.org/lists.html#faq

Re: error processing 220-noteGNUstack-01.patch

[Yann E. MORIN]

Hello all!

On Wednesday 29 April 2009 22:59:25 ng@piments.com wrote:
> Yann E. MORIN wrote:
> > OK. I guess reiser4 is case-sensitive, is it not?
> Sure, it's case sensitive.
> 
> but this does not seem to be a false error , the file does already 
> exist.

No, it does (should) not exist. The patch creates the file:
  boehm-gc/ia64_save_regs_in_stack.S

and the file that already exists is:
  boehm-gc/ia64_save_regs_in_stack.s

The difference is subtle: the existing file has a lower-case 's' as
extension, while the file that gets created has a upper-case 'S'.

The patch does apply cleanly here on ext3.

Hence the question about the case-sensitivity of your file system.
If your file system is case sensitive, then if a file named "foo"
exists, then testing for existence of "FOO" will succeed. And the
patch utility will fail.

This happens all the time under Windows/Cygwin.

> I have not rerun anything since the run that failed. The error it  
> states is correct.

Telling it again and again will not make it true:
 that file - does - not - exist - in the gcc-4.3.2 tarball.

If you have such a file, then:
- either your file system is not case-sensitive,
- or your gcc-4.3.2 tarball is not pristine.

> The file exists and has the expected content. Why is  
> the patch being called?

The patch creates the .S (upper-case) file and removes a file with a
.s (lower-case) extension, and the same content. 

> Is there a way with ct-ng to remove this patch and rebuild without it 
> being refeched?

No, there is none.

> Or can you suggest another work around?

Let's try to understand what on Earth is going on:
- if there is a bug, let's shoot it down.
- if there is an operator error, let's update the doc.

Don't workaround bugs/errors: they're still there lurking until we
forget about the workaround, and we'll get bitten back sooner or later.

Regards,
Yann E. MORIN.

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +0/33 662376056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| --==< ^_^ >==-- `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
`------------------------------^-------^------------------^--------------------'


--
For unsubscribe information see http://sourceware.org/lists.html#faq

Re: error processing 220-noteGNUstack-01.patch

[ng]

Yann E. MORIN wrote:
> Hello all!
> 
> On Wednesday 29 April 2009 22:59:25 ng@piments.com wrote:
>> Yann E. MORIN wrote:
>>> OK. I guess reiser4 is case-sensitive, is it not?
>> Sure, it's case sensitive.
>>
>> but this does not seem to be a false error , the file does already 
>> exist.
> 
> No, it does (should) not exist. The patch creates the file:
>   boehm-gc/ia64_save_regs_in_stack.S
> 
> and the file that already exists is:
>   boehm-gc/ia64_save_regs_in_stack.s
> 
> The difference is subtle: the existing file has a lower-case 's' as
> extension, while the file that gets created has a upper-case 'S'.
> 
> The patch does apply cleanly here on ext3.
> 
> Hence the question about the case-sensitivity of your file system.
> If your file system is case sensitive, then if a file named "foo"
> exists, then testing for existence of "FOO" will succeed. And the
> patch utility will fail.
> 
> This happens all the time under Windows/Cygwin.
> 
>> I have not rerun anything since the run that failed. The error it  
>> states is correct.
> 
> Telling it again and again will not make it true:
>  that file - does - not - exist - in the gcc-4.3.2 tarball.
> 
> If you have such a file, then:
> - either your file system is not case-sensitive,
> - or your gcc-4.3.2 tarball is not pristine.
> 
>> The file exists and has the expected content. Why is  
>> the patch being called?
> 
> The patch creates the .S (upper-case) file and removes a file with a
> .s (lower-case) extension, and the same content. 
> 
>> Is there a way with ct-ng to remove this patch and rebuild without it 
>> being refeched?
> 
> No, there is none.
> 
>> Or can you suggest another work around?
> 
> Let's try to understand what on Earth is going on:
> - if there is a bug, let's shoot it down.
> - if there is an operator error, let's update the doc.
> 
> Don't workaround bugs/errors: they're still there lurking until we
> forget about the workaround, and we'll get bitten back sooner or later.
> 
> Regards,
> Yann E. MORIN.
> 
OK, I just installed a clean 1.4.0 and loaded arm-unknown-linux-gnuabi 
sample which included the addons and cvs fetch. The extra options are 
now showing in C library menu.

I modified ARCH to armv4t to match my hardware and the gcc and glibc 
issues have disappeared.

I can only assume that the problems I was having came from installing 
1.4.0 on top of 1.3.4 rather than a clean installation and/or some other 
cruft.

I have not tested the toolchain yet but I got a clean build.

Thanks for your help, Yann.

regards.







--
For unsubscribe information see http://sourceware.org/lists.html#faq