Re: sbuild race with sporadic build breakages

[Jan Kiszka <jan.kiszka@siemens.com>]

On 24.05.24 11:30, Anton Mikanovich wrote:
> 24/05/2024 11:52, Jan Kiszka wrote:
>> So, the variable may actually also not be evaluated per task but at some
>> point centrally? Then this is even more broken. Let's just use something
>> like a task hash here, stable and still unique inside a run.
>>
>> Jan
>>
> Sbuild is getting SBUILD_CHROOT value only once via command line option.
> And it is already calculated and expanded at this point.

And that is where I'm not so sure, given the problems we see.

> So no meter which keys will be used for SBUILD_CHROOT path the issue will
> probably be there: tasks paths still will be different and sbuild still
> will
> failing in case getting the path value of other wrong task.
> 

Why speaks against this?

diff --git a/meta/classes/sbuild.bbclass b/meta/classes/sbuild.bbclass
index f1193c20..3f27a2ef 100644
--- a/meta/classes/sbuild.bbclass
+++ b/meta/classes/sbuild.bbclass
@@ -7,7 +7,7 @@ SCHROOT_MOUNTS ?= ""
 
 inherit crossvars
 
-SBUILD_CHROOT ?= "${DEBDISTRONAME}-${SCHROOT_USER}-${ISAR_BUILD_UUID}-${@os.getpid()}"
+SBUILD_CHROOT ?= "${DEBDISTRONAME}-${SCHROOT_USER}-${ISAR_BUILD_UUID}-${BB_TASKHASH}"
 
 SBUILD_CONF_DIR ?= "${SCHROOT_CONF}/${SBUILD_CHROOT}"
 SCHROOT_CONF_FILE ?= "${SCHROOT_CONF}/chroot.d/${SBUILD_CHROOT}"


We definitely need to get rid of that getpid.

Jan

-- 
Siemens AG, Technology
Linux Expert Center