Re: [PATCH] fix the build issue of ppp

[Joshua Lock <josh@linux.intel.com>]

On 05/06/12 12:17, Khem Raj wrote:
> On Tue, Jun 5, 2012 at 11:59 AM, Joshua Lock<josh@linux.intel.com>  wrote:
>> On 04/06/12 20:32, b19537@freescale.com wrote:
>>>
>>> From: Zhenhua Luo<b19537@freescale.com>
>>>
>>>      this patch is for edison branch of poky
>>>
>>>      fix following build issue of ppp:
>>>          make[2]: Entering directory
>>> `/local/home/b19537/workspace/sdk-devel/build_p4080ds_release/tmp/work/ppce500mc-fsl-linux/ppp-2.4.5-r2/ppp-2.4.5/pppd/plugins/rp-pppoe'
>>>          powerpc-fsl-linux-gcc  -m32 -mhard-float  -mcpu=e500mc  -Wl,-O1
>>> -Wl,--hash-style=gnu -Wl,--as-needed
>>> --sysroot=/local/home/b19537/workspace/sdk-devel/build_p4080ds_release/tmp/sysroots/p4080ds
>>> -O2 -g -I../../../include '-DRP_VERSION="3.8p"' -I../../.. -c -o plugin.o
>>> -fPIC plugin.c
>>>          In file included from plugin.c:53:0:
>>>
>>> /local/home/b19537/workspace/sdk-devel/build_p4080ds_release/tmp/sysroots/p4080ds/usr/include/linux/if_pppox.h:85:25:
>>> error: field 'pppol2tp' has incomplete type
>>>          make[2]: *** [plugin.o] Error 1
>>>          make[2]: *** Waiting for unfinished jobs....
>>>          make[2]: Leaving directory
>>> `/local/home/b19537/workspace/sdk-devel/build_p4080ds_release/tmp/work/ppce500mc-fsl-linux/ppp-2.4.5-r2/ppp-2.4.5/pppd/plugins/rp-pppoe'
>>
>>
>> PPP builds fine here, can you share some more information such as what
>> causes the issue and where this patch has been tested?
>>
>
> Did you build it for ppc ?

I did, yes.

> ppc captures its own version of linux/if_pppol2tp.h and if this is not in sync
> with kernel bad things can happen on different architectures. I think in case
> of ppc here some reference is coming in and including this file instead of the
> one from kernel-headers for ppc but not for may be x86. This patch is safe
> but it would be better if whole capturing of this file can be avoided.

That was my assumption but with very little information provided I'm 
reluctant to take the patch.

For example if the kernel with which the issue was triggered is provided 
by a separate layer it seems like this patch belongs in a bbappend in 
the same layer, no?

If I can build for qemu machines without this patch will not the 
application of the patch have a negative effect?

Sure, I can likely answer each of these questions myself if I spend some 
CPU cycles running builds but

>
>> Further, there's no header on the patch. Please see the guidelines at:
>>
>> https://wiki.yoctoproject.org/wiki/Contribution_Guidelines
>>
>
> yes thats needed for better patch management
>> Cheers,
>> Joshua
>>
>>
>>>
>>> Signed-off-by: Zhenhua Luo<b19537@freescale.com>
>>> ---
>>>   .../ppp-fix-build-issue-of-rp-pppoe.so.patch       |   23
>>> ++++++++++++++++++++
>>>   meta/recipes-connectivity/ppp/ppp_2.4.5.bb         |    3 +-
>>>   2 files changed, 25 insertions(+), 1 deletions(-)
>>>   create mode 100644
>>> meta/recipes-connectivity/ppp/ppp-2.4.5/ppp-fix-build-issue-of-rp-pppoe.so.patch
>>>
>>> diff --git
>>> a/meta/recipes-connectivity/ppp/ppp-2.4.5/ppp-fix-build-issue-of-rp-pppoe.so.patch
>>> b/meta/recipes-connectivity/ppp/ppp-2.4.5/ppp-fix-build-issue-of-rp-pppoe.so.patch
>>> new file mode 100644
>>> index 0000000..ed2141a
>>> --- /dev/null
>>> +++
>>> b/meta/recipes-connectivity/ppp/ppp-2.4.5/ppp-fix-build-issue-of-rp-pppoe.so.patch
>>> @@ -0,0 +1,23 @@
>>> +--- a/include/linux/if_pppol2tp.h      2012-06-04 04:22:37.966729643
>>> -0500
>>> ++++ b/include/linux/if_pppol2tp.h      2012-06-04 04:31:04.325905458
>>> -0500
>>> +@@ -32,6 +32,20 @@
>>> +       __u16 d_tunnel, d_session;      /* For sending outgoing packets */
>>> + };
>>> +
>>> ++/* The L2TPv3 protocol changes tunnel and session ids from 16 to 32
>>> ++ * bits. So we need a different sockaddr structure.
>>> ++ */
>>> ++struct pppol2tpv3_addr {
>>> ++       pid_t   pid;                    /* pid that owns the fd.
>>> ++                                        * 0 =>    current */
>>> ++       int     fd;                     /* FD of UDP or IP socket to use
>>> */
>>> ++
>>> ++       struct sockaddr_in addr;        /* IP address and port to send to
>>> */
>>> ++
>>> ++       __u32 s_tunnel, s_session;      /* For matching incoming packets
>>> */
>>> ++       __u32 d_tunnel, d_session;      /* For sending outgoing packets
>>> */
>>> ++};
>>> ++
>>> + /* Socket options:
>>> +  * DEBUG      - bitmask of debug message categories
>>> +  * SENDSEQ    - 0 =>    don't send packets with sequence numbers
>>> diff --git a/meta/recipes-connectivity/ppp/ppp_2.4.5.bb
>>> b/meta/recipes-connectivity/ppp/ppp_2.4.5.bb
>>> index 72464bf..6c4bd08 100644
>>> --- a/meta/recipes-connectivity/ppp/ppp_2.4.5.bb
>>> +++ b/meta/recipes-connectivity/ppp/ppp_2.4.5.bb
>>> @@ -10,7 +10,7 @@ LIC_FILES_CHKSUM =
>>> "file://pppd/ccp.c;beginline=1;endline=29;md5=e2c43fe6e81ff77
>>>
>>>   file://pppd/plugins/passprompt.c;beginline=1;endline=10;md5=3bcbcdbf0e369c9a3e0b8c8275b065d8
>>> \
>>>
>>>   file://pppd/tdb.c;beginline=1;endline=27;md5=4ca3a9991b011038d085d6675ae7c4e6
>>> \
>>>
>>>   file://chat/chat.c;beginline=1;endline=15;md5=0d374b8545ee5c62d7aff1acbd38add2"
>>> -PR = "r2"
>>> +PR = "r3"
>>>
>>>   SRC_URI = "http://ppp.samba.org/ftp/ppp/ppp-${PV}.tar.gz \
>>>              file://makefile.patch \
>>> @@ -18,6 +18,7 @@ SRC_URI = "http://ppp.samba.org/ftp/ppp/ppp-${PV}.tar.gz
>>> \
>>>              file://pppd-resolv-varrun.patch \
>>>              file://enable-ipv6.patch \
>>>              file://makefile-remove-hard-usr-reference.patch \
>>> +           file://ppp-fix-build-issue-of-rp-pppoe.so.patch \
>>>              file://pon \
>>>              file://poff \
>>>              file://init \
>>
>>
>> --
>> Joshua Lock
>>         Yocto Project
>>         Intel Open Source Technology Centre
>>
>> _______________________________________________
>> yocto mailing list
>> yocto@yoctoproject.org
>> https://lists.yoctoproject.org/listinfo/yocto

-- 
Joshua Lock
         Yocto Project
         Intel Open Source Technology Centre