Re: [meta-arm] [PATCH v2 3/6] arm/optee-os: deploy tee.elf for u-boot

["Denys Dmytriyenko" <denis@denix.org>]

On Tue, Apr 20, 2021 at 11:35:55AM +0200, Yann Dirson wrote:
> *
> 
> Le mar. 20 avr. 2021 à 09:54, Yann Dirson
> <yann.dirson@blade-group.com> a écrit :
> >
> > Le mar. 20 avr. 2021 à 02:18, Denys Dmytriyenko <denis@denix.org> a écrit :
> > >
> > > On Mon, Apr 19, 2021 at 05:34:13PM +0100, Ross Burton wrote:
> > > > This is causing the CI to break:
> > > >
> > > > ERROR: optee-os-3.11.0-r0 do_package_qa: QA Issue: optee-os: ELF
> > > > binary /lib/firmware/tee.elf has relocations in .text [textrel]
> > > >
> > > > Can you either fix or add an INSANE_SKIP with a comment?
> > >
> > > Ah, sorry, forgot to mention that textrel needs to be added to INSANE_SKIP now
> > > that tee.elf gets packaged.
> >
> > I'm surprised, how is it that this test is not triggered on all boxes
> > (ie. on the master CI but not on testing
> > or on my box) ?  I'm all for submitting a fix but it's harder to be
> > correct why I cannot reproduce the problem :)
> 
> Digging this I am surprised by how hard it is to force a package_qa rerun:
> - the package_qa_* functions text is not included, so even changing
> them does not
>   cause them to re-run
> - WARN_QA and ERROR_QA are part of BB_HASHBASE_WHITELIST, so adjusting
>   this does not cause a rerun either
> 
> It would seem interesting to have all of this part of the package_qa
> siginfo, right ?
> 
> I have the clear feeling the workflow to investigate the issue is
> really suboptimal
> ("bitbake -c clean && find sstate-cache -delete" between each run).

I usually do -c cleansstate and then force -c compile -f to invalidate sstate 
for a package.


> As for the issue at hand: I can now trigger package_qa at will, and
> adding more traces
> in textrel test I can see it is really executed on tee.elf, but it
> just passes the test.
> 
> "objdump -p" and "objdump -r" just show nothing, so indeed no TEXTREL.
> 
> On what configuration does the CI get this ?

I wonder if tee.elf for your platform is different. Are you building for qemu 
or a real machine? I saw this QA issue reported for TI platforms before.


> > > > Ross
> > > >
> > > > On Wed, 14 Apr 2021 at 14:48, Yann Dirson <yann.dirson@blade-group.com> wrote:
> > > > >
> > > > > From: Yann Dirson <yann@blade-group.com>
> > > > >
> > > > > Integration of OP-TEE in the u-boot image is done with `TEE=tee.elf`,
> > > > > so this build artifact must be exported.
> > > > >
> > > > > U-boot integration can now be done in a BSP layer with something like
> > > > > this in `u-boot_%.bbappend`:
> > > > >
> > > > >  # OP-TEE, dependent on "optee" DISTRO_FEATURES
> > > > >
> > > > >  OPTEE_OEMAKE ?= " TEE=${DEPLOY_DIR_IMAGE}/optee/tee.elf"
> > > > >
> > > > >  EXTRA_OEMAKE_append = " ${PACKAGECONFIG_CONFARGS}"
> > > > >  PACKAGECONFIG[optee] = "${OPTEE_OEMAKE},,optee-os"
> > > > >  PACKAGECONFIG_append = " ${@bb.utils.filter('DISTRO_FEATURES', 'optee', d)}"
> > > > >
> > > > > Signed-off-by: Yann Dirson <yann@blade-group.com>
> > > > > ---
> > > > >  meta-arm/recipes-security/optee/optee-os.inc | 2 +-
> > > > >  1 file changed, 1 insertion(+), 1 deletion(-)
> > > > >
> > > > > diff --git a/meta-arm/recipes-security/optee/optee-os.inc b/meta-arm/recipes-security/optee/optee-os.inc
> > > > > index 34f586a..08cc2f0 100644
> > > > > --- a/meta-arm/recipes-security/optee/optee-os.inc
> > > > > +++ b/meta-arm/recipes-security/optee/optee-os.inc
> > > > > @@ -45,7 +45,7 @@ do_compile[cleandirs] = "${B}"
> > > > >  do_install() {
> > > > >      #install core in firmware
> > > > >      install -d ${D}${nonarch_base_libdir}/firmware/
> > > > > -    install -m 644 ${B}/core/*.bin ${D}${nonarch_base_libdir}/firmware/
> > > > > +    install -m 644 ${B}/core/*.bin ${B}/core/tee.elf ${D}${nonarch_base_libdir}/firmware/
> > > > >
> > > > >      #install TA devkit
> > > > >      install -d ${D}${includedir}/optee/export-user_ta/
> > > > > --
> > > > > 2.30.2
> > > > >
> > > > >
> > > > >
> > > > >
> > >
> > > --
> > > Regards,
> > > Denys Dmytriyenko <denis@denix.org>
> > > PGP: 0x420902729A92C964 - https://denix.org/0x420902729A92C964
> > > Fingerprint: 25FC E4A5 8A72 2F69 1186  6D76 4209 0272 9A92 C964
> >
> >
> >
> > --
> > Yann Dirson <yann@blade-group.com>
> > Blade / Shadow -- http://shadow.tech
> 
> 
> 
> -- 
> Yann Dirson <yann@blade-group.com>
> Blade / Shadow -- http://shadow.tech
> 

-- 
Regards,
Denys Dmytriyenko <denis@denix.org>
PGP: 0x420902729A92C964 - https://denix.org/0x420902729A92C964
Fingerprint: 25FC E4A5 8A72 2F69 1186  6D76 4209 0272 9A92 C964