[dpdk-ci] UNH-IOL ABI Failures

[dpdk-ci] UNH-IOL ABI Failures

[Brandon Lo]

Hi all,

We have noticed some failures occurring repeatedly in our CI system
regarding the ABI test.

I looked into the failures and saw that there were some failures
caused by a clock skew issue on the ARM ABI test.
Those failures have been rerun and the report will be resent to
patchworks for the updated result.

However, the earlier test runs seem to report an actual failure:

"1 function with some indirect sub-type change:

  [C]'function rte_security_session*
rte_security_session_create(rte_security_ctx*,
rte_security_session_conf*, rte_mempool*, rte_mempool*)' at
rte_security.c:43:1 has some indirect sub-type changes:
    parameter 2 of type 'rte_security_session_conf*' has sub-type changes:
      in pointed to type 'struct rte_security_session_conf' at
rte_security.h:366:1:
        type size hasn't changed
        1 data member change:
         type of 'rte_crypto_sym_xform*
rte_security_session_conf::crypto_xform' changed:
           in pointed to type 'struct rte_crypto_sym_xform' at
rte_crypto_sym.h:575:1:
             type size hasn't changed
             1 data member changes (1 filtered):

Error: ABI issue reported for 'abidiff --suppr
dpdk/devtools/libabigail.abignore --no-added-syms --headers-dir1
reference/usr/local/include --headers-dir2 build/usr/local/include
reference/dump/librte_security.dump build/dump/librte_security.dump'

ABIDIFF_ABI_CHANGE, this change requires a review (abidiff flagged
this as a potential issue)."

Can anyone please confirm that this is a real ABI issue or if this
requires a change on the UNH-IOL side.

Thanks,
Brandon

-- 
Brandon Lo

UNH InterOperability Laboratory

21 Madbury Rd, Suite 100, Durham, NH 03824

blo@iol.unh.edu

www.iol.unh.edu

Re: [dpdk-ci] UNH-IOL ABI Failures

[Thomas Monjalon]

20/04/2021 17:30, Brandon Lo:
> Hi all,
> 
> We have noticed some failures occurring repeatedly in our CI system
> regarding the ABI test.
> 
> I looked into the failures and saw that there were some failures
> caused by a clock skew issue on the ARM ABI test.
> Those failures have been rerun and the report will be resent to
> patchworks for the updated result.
> 
> However, the earlier test runs seem to report an actual failure:
> 
> "1 function with some indirect sub-type change:
> 
>   [C]'function rte_security_session*
> rte_security_session_create(rte_security_ctx*,
> rte_security_session_conf*, rte_mempool*, rte_mempool*)' at
> rte_security.c:43:1 has some indirect sub-type changes:
>     parameter 2 of type 'rte_security_session_conf*' has sub-type changes:
>       in pointed to type 'struct rte_security_session_conf' at
> rte_security.h:366:1:
>         type size hasn't changed
>         1 data member change:
>          type of 'rte_crypto_sym_xform*
> rte_security_session_conf::crypto_xform' changed:
>            in pointed to type 'struct rte_crypto_sym_xform' at
> rte_crypto_sym.h:575:1:
>              type size hasn't changed
>              1 data member changes (1 filtered):
> 
> Error: ABI issue reported for 'abidiff --suppr
> dpdk/devtools/libabigail.abignore --no-added-syms --headers-dir1
> reference/usr/local/include --headers-dir2 build/usr/local/include
> reference/dump/librte_security.dump build/dump/librte_security.dump'
> 
> ABIDIFF_ABI_CHANGE, this change requires a review (abidiff flagged
> this as a potential issue)."
> 
> Can anyone please confirm that this is a real ABI issue or if this
> requires a change on the UNH-IOL side.

It requires libabigail 1.8 on UNH-IOL side.

Re: [dpdk-ci] UNH-IOL ABI Failures

[Brandon Lo]

Okay, I am updating the ABI version on the machines and rerunning the
affected patches.
I had mistakenly installed libabigail 1.6 since that was the default
in the linux-build.sh script.

Thanks,
Brandon

On Tue, Apr 20, 2021 at 11:37 AM Thomas Monjalon <thomas@monjalon.net> wrote:
>
> 20/04/2021 17:30, Brandon Lo:
> > Hi all,
> >
> > We have noticed some failures occurring repeatedly in our CI system
> > regarding the ABI test.
> >
> > I looked into the failures and saw that there were some failures
> > caused by a clock skew issue on the ARM ABI test.
> > Those failures have been rerun and the report will be resent to
> > patchworks for the updated result.
> >
> > However, the earlier test runs seem to report an actual failure:
> >
> > "1 function with some indirect sub-type change:
> >
> >   [C]'function rte_security_session*
> > rte_security_session_create(rte_security_ctx*,
> > rte_security_session_conf*, rte_mempool*, rte_mempool*)' at
> > rte_security.c:43:1 has some indirect sub-type changes:
> >     parameter 2 of type 'rte_security_session_conf*' has sub-type changes:
> >       in pointed to type 'struct rte_security_session_conf' at
> > rte_security.h:366:1:
> >         type size hasn't changed
> >         1 data member change:
> >          type of 'rte_crypto_sym_xform*
> > rte_security_session_conf::crypto_xform' changed:
> >            in pointed to type 'struct rte_crypto_sym_xform' at
> > rte_crypto_sym.h:575:1:
> >              type size hasn't changed
> >              1 data member changes (1 filtered):
> >
> > Error: ABI issue reported for 'abidiff --suppr
> > dpdk/devtools/libabigail.abignore --no-added-syms --headers-dir1
> > reference/usr/local/include --headers-dir2 build/usr/local/include
> > reference/dump/librte_security.dump build/dump/librte_security.dump'
> >
> > ABIDIFF_ABI_CHANGE, this change requires a review (abidiff flagged
> > this as a potential issue)."
> >
> > Can anyone please confirm that this is a real ABI issue or if this
> > requires a change on the UNH-IOL side.
>
> It requires libabigail 1.8 on UNH-IOL side.
>
>
>


-- 

Brandon Lo

UNH InterOperability Laboratory

21 Madbury Rd, Suite 100, Durham, NH 03824

blo@iol.unh.edu

www.iol.unh.edu