All of lore.kernel.org
 help / color / mirror / Atom feed
* Experimental_DCCP_source_tree: Snapshot patch for kernel 2.6.24
@ 2008-05-26 11:53 Giuseppe Galeota
  2008-05-26 12:11 ` Gerrit Renker
                   ` (3 more replies)
  0 siblings, 4 replies; 5+ messages in thread
From: Giuseppe Galeota @ 2008-05-26 11:53 UTC (permalink / raw)
  To: dccp

Hi,
I have followed the guide from 
http://www.linux-foundation.org/en/Net:DCCP_Testing in order to use the 
Monitoring DCCP.
I also need to apply the web100 patch, that is valid only for a kernel 
2.6.*24* (http://www.web100.org).

So, I have downloaded the kernel 2.6.24 from 
http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.24.tar.gz, and I 
applied the web100 patch (that is, 
web100-2.6.24-2.5.19-200802211940.patch , downloaded from 
http://www.web100.org). This is OK.
After that, I tried to enable the Experimental_DCCP_source_tree using 
the Snapshot patch, that I have downloaded from 
http://www.erg.abdn.ac.uk/users/gerrit/dccp/testing_dccp/test-tree/test-tree_2.6.24.diff.gz, 
but  without success.
(If you want, I can post the result).

I also tried to apply the web100 patch to the kernel ( or nigthly 
snapshot of the DCCP tree ) downloaded from 
http://eden-feed.erg.abdn.ac.uk/latest-dccp-test-tree.tar.bz2, but 
without success.

Have you suggestions to resolve this situation?

Giuseppe Galeota

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: Experimental_DCCP_source_tree: Snapshot patch for kernel 2.6.24
  2008-05-26 11:53 Experimental_DCCP_source_tree: Snapshot patch for kernel 2.6.24 Giuseppe Galeota
@ 2008-05-26 12:11 ` Gerrit Renker
  2008-05-27 10:23 ` Gerrit Renker
                   ` (2 subsequent siblings)
  3 siblings, 0 replies; 5+ messages in thread
From: Gerrit Renker @ 2008-05-26 12:11 UTC (permalink / raw)
  To: dccp

Hi Giuseppe,

you are the second one to report this and I am beginning to think that
the snapshot patches may not be such a good idea. The problem is that
they are taken on the day that the `EXTRAVERSION' variable changes in
the top-level Makefile.

I have downloaded a 2.6.24, and found that 
http://www.erg.abdn.ac.uk/users/gerrit/dccp/testing_dccp/test-tree/test-tree_2.6.24-rc8.diff.gz

applied cleanly, i.e. probably this is the version later published as
2.6.24 on kernel.org. Could you give this a try please?

But even then, the older patches are lacking some of the more recent
fixes. I will try to backport the latest snapshot patch to 2.6.24 if
I can find some time.

Gerrit

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: Experimental_DCCP_source_tree: Snapshot patch for kernel 2.6.24
  2008-05-26 11:53 Experimental_DCCP_source_tree: Snapshot patch for kernel 2.6.24 Giuseppe Galeota
  2008-05-26 12:11 ` Gerrit Renker
@ 2008-05-27 10:23 ` Gerrit Renker
  2008-05-27 19:26 ` Giuseppe Galeota
  2008-05-28  6:55 ` Gerrit Renker
  3 siblings, 0 replies; 5+ messages in thread
From: Gerrit Renker @ 2008-05-27 10:23 UTC (permalink / raw)
  To: dccp

> However, I tried to compile the new kernel (importing my old file .config), but I could see some errors about net/dccp/feat.
> (I have enclosed a part of the result).
> What do you think about this?
>
This is what I meant with not having the latest fixes. The problem is
because you are compiling your kernel with the CONFIG_IP_DCCP_xxx_DEBUG
turned off. The bug had been pointed out last week by Wei Yongjun and
has been fixed since in the test tree, but not in the older patches.

I will fix the 2.6.24/25 patches also, you can fix this by either
 * compiling the kernel with debug options for DCCP turned on
  (and then I am sure it works);
 * change the following hunk in net/dccp/feat.c

from:

	#else  /* ! CONFIG_IP_DCCP_DEBUG */
	#define dccp_feat_print_opt(opt, feat, fval, mandatory)
	#define dccp_feat_print_fnlist(fn_list)
	#endif

to:
	#else  /* ! CONFIG_IP_DCCP_DEBUG */
	#define dccp_feat_print_opt(opt, feat, val, len, mandatory)
	#define dccp_feat_print_fnlist(fn_list)
	#endif

Any name for the arguments will do - it is just an empty declaration.

Thanks
Gerrit

| net/dccp/feat.c:546:54: error: macro "dccp_feat_print_opt" passed 5 arguments, but takes just 4
| net/dccp/feat.c: In function ???dccp_feat_insert_opts???:
| net/dccp/feat.c:546: error: ???dccp_feat_print_opt??? undeclared (first use in this function)
| net/dccp/feat.c:546: error: (Each undeclared identifier is reported only once
| net/dccp/feat.c:546: error: for each function it appears in.)
| net/dccp/feat.c:962:55: error: macro "dccp_feat_print_opt" passed 5 arguments, but takes just 4
| net/dccp/feat.c: In function ???dccp_feat_change_recv???:
| net/dccp/feat.c:962: error: ???dccp_feat_print_opt??? undeclared (first use in this function)
| net/dccp/feat.c:1071:55: error: macro "dccp_feat_print_opt" passed 5 arguments, but takes just 4
| net/dccp/feat.c: In function ???dccp_feat_confirm_recv???:
| net/dccp/feat.c:1071: error: ???dccp_feat_print_opt??? undeclared (first use in this function)
| net/dccp/feat.c:1168:52: error: macro "dccp_feat_print_opt" passed 5 arguments, but takes just 4
| net/dccp/feat.c: In function ???dccp_feat_handle_nn_established???:
| net/dccp/feat.c:1168: error: ???dccp_feat_print_opt??? undeclared (first use in this function)
| make[3]: *** [net/dccp/feat.o] Error 1
| make[2]: *** [net/dccp] Error 2
| make[1]: *** [net] Error 2
| make[1]: Leaving directory `/usr/src/linux-source-2.6.24-org'
| make: *** [debian/stamp-build-kernel] Error 2
|


--

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: Experimental_DCCP_source_tree: Snapshot patch for kernel 2.6.24
  2008-05-26 11:53 Experimental_DCCP_source_tree: Snapshot patch for kernel 2.6.24 Giuseppe Galeota
  2008-05-26 12:11 ` Gerrit Renker
  2008-05-27 10:23 ` Gerrit Renker
@ 2008-05-27 19:26 ` Giuseppe Galeota
  2008-05-28  6:55 ` Gerrit Renker
  3 siblings, 0 replies; 5+ messages in thread
From: Giuseppe Galeota @ 2008-05-27 19:26 UTC (permalink / raw)
  To: dccp

Dear Gerrit, 

I made this:
 1: I have turned on DCCP Protocol-->Dccp CCIDs Configurations-->CCIS2/CCID3 debugging messages in the file .config(using make gconfig)
 2: I have changed the following hunk in net/dccp/feat.c, as you said
from:

	#else  /* ! CONFIG_IP_DCCP_DEBUG */
	#define dccp_feat_print_opt(opt, feat, fval, mandatory)
	#define dccp_feat_print_fnlist(fn_list)
	#endif

to:
	#else  /* ! CONFIG_IP_DCCP_DEBUG */
	#define dccp_feat_print_opt(opt, feat, val, len, mandatory)
	#define dccp_feat_print_fnlist(fn_list)
	#endif

Than, I have compiled the kernel, and regarded /net/dccp/feat.c I only read a warning:
  /net/dccp/feat.c: .........warning: dccp_feat_oname: defined but not used

So, I think that is all OK, isn't it?

Thanks
Giuseppe




Gerrit Renker ha scritto:
>> However, I tried to compile the new kernel (importing my old file .config), but I could see some errors about net/dccp/feat.
>> (I have enclosed a part of the result).
>> What do you think about this?
>>
>>     
> This is what I meant with not having the latest fixes. The problem is
> because you are compiling your kernel with the CONFIG_IP_DCCP_xxx_DEBUG
> turned off. The bug had been pointed out last week by Wei Yongjun and
> has been fixed since in the test tree, but not in the older patches.
>
> I will fix the 2.6.24/25 patches also, you can fix this by either
>  * compiling the kernel with debug options for DCCP turned on
>   (and then I am sure it works);
>  * change the following hunk in net/dccp/feat.c
>
> from:
>
> 	#else  /* ! CONFIG_IP_DCCP_DEBUG */
> 	#define dccp_feat_print_opt(opt, feat, fval, mandatory)
> 	#define dccp_feat_print_fnlist(fn_list)
> 	#endif
>
> to:
> 	#else  /* ! CONFIG_IP_DCCP_DEBUG */
> 	#define dccp_feat_print_opt(opt, feat, val, len, mandatory)
> 	#define dccp_feat_print_fnlist(fn_list)
> 	#endif
>
> Any name for the arguments will do - it is just an empty declaration.
>
> Thanks
> Gerrit
>
> | net/dccp/feat.c:546:54: error: macro "dccp_feat_print_opt" passed 5 arguments, but takes just 4
> | net/dccp/feat.c: In function ???dccp_feat_insert_opts???:
> | net/dccp/feat.c:546: error: ???dccp_feat_print_opt??? undeclared (first use in this function)
> | net/dccp/feat.c:546: error: (Each undeclared identifier is reported only once
> | net/dccp/feat.c:546: error: for each function it appears in.)
> | net/dccp/feat.c:962:55: error: macro "dccp_feat_print_opt" passed 5 arguments, but takes just 4
> | net/dccp/feat.c: In function ???dccp_feat_change_recv???:
> | net/dccp/feat.c:962: error: ???dccp_feat_print_opt??? undeclared (first use in this function)
> | net/dccp/feat.c:1071:55: error: macro "dccp_feat_print_opt" passed 5 arguments, but takes just 4
> | net/dccp/feat.c: In function ???dccp_feat_confirm_recv???:
> | net/dccp/feat.c:1071: error: ???dccp_feat_print_opt??? undeclared (first use in this function)
> | net/dccp/feat.c:1168:52: error: macro "dccp_feat_print_opt" passed 5 arguments, but takes just 4
> | net/dccp/feat.c: In function ???dccp_feat_handle_nn_established???:
> | net/dccp/feat.c:1168: error: ???dccp_feat_print_opt??? undeclared (first use in this function)
> | make[3]: *** [net/dccp/feat.o] Error 1
> | make[2]: *** [net/dccp] Error 2
> | make[1]: *** [net] Error 2
> | make[1]: Leaving directory `/usr/src/linux-source-2.6.24-org'
> | make: *** [debian/stamp-build-kernel] Error 2
> |
>
>
> --
> .
>
>   

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: Experimental_DCCP_source_tree: Snapshot patch for kernel 2.6.24
  2008-05-26 11:53 Experimental_DCCP_source_tree: Snapshot patch for kernel 2.6.24 Giuseppe Galeota
                   ` (2 preceding siblings ...)
  2008-05-27 19:26 ` Giuseppe Galeota
@ 2008-05-28  6:55 ` Gerrit Renker
  3 siblings, 0 replies; 5+ messages in thread
From: Gerrit Renker @ 2008-05-28  6:55 UTC (permalink / raw)
  To: dccp

Giuseppe, -
| 
|  Than, I have compiled the kernel, and regarded /net/dccp/feat.c I only read a warning:
|    /net/dccp/feat.c: .........warning: dccp_feat_oname: defined but not used
| 
|  So, I think that is all OK, isn't it?
| 
Yes, it should work now. Typically each patch that is in the directory
has been tested to work here on 4..7 different computers, but it seems not
sufficiently with all combinations of options.

The warning message will disappear when enabling 'DCCP debug messages'
(via the 'DCCP kernel hacking' submenu), but you are right, it shouldn't
happen in any combination of configuration options.

I will fix this in the test tree and also in the 2.6.24 snapshot patch.

Thanks a lot for reporting this,
Gerrit

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2008-05-28  6:55 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-05-26 11:53 Experimental_DCCP_source_tree: Snapshot patch for kernel 2.6.24 Giuseppe Galeota
2008-05-26 12:11 ` Gerrit Renker
2008-05-27 10:23 ` Gerrit Renker
2008-05-27 19:26 ` Giuseppe Galeota
2008-05-28  6:55 ` Gerrit Renker

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.