Re: [PATCH v2] Set STAMPCLEAN variable

[Claudius Heine <claudius.heine.ext@siemens.com>]

On 13/08/2019 10.01, Jan Kiszka wrote:
> On 13.08.19 09:27, Claudius Heine wrote:
>> Hi Jan,
>>
>> On 12/08/2019 19.15, [ext] Jan Kiszka wrote:
>>> From: Jan Kiszka <jan.kiszka@siemens.com>
>>>
>>> This ensures that stamp files are cleaned when a task is rebuilt. Not
>>> providing this pattern likely caused a lot of failures when doing
>>> partial rebuilds.
>>>
>>> As we are at it, switch do_clean to STAMPCLEAN as well.
>>>
>>> Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
>>> ---
>>>   meta/classes/base.bbclass | 4 ++--
>>>   meta/conf/bitbake.conf    | 1 +
>>>   2 files changed, 3 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/meta/classes/base.bbclass b/meta/classes/base.bbclass
>>> index ca80bf8..b6638ca 100644
>>> --- a/meta/classes/base.bbclass
>>> +++ b/meta/classes/base.bbclass
>>> @@ -206,8 +206,8 @@ python do_clean() {
>>>       workdir = d.expand("${WORKDIR}")
>>>       subprocess.check_call(["sudo", "rm", "-rf", workdir])
>>> -    stamppath = bb.data.expand(d.getVar('STAMP', False), d)
>>> -    stampdirs = glob.glob(stamppath + ".*")
>>> +    stampclean = bb.data.expand(d.getVar('STAMPCLEAN', False), d)
>>> +    stampdirs = glob.glob(stampclean)
>>>       subprocess.check_call(["sudo", "rm", "-rf"] + stampdirs)
>>>   }
>>> diff --git a/meta/conf/bitbake.conf b/meta/conf/bitbake.conf
>>> index 3d18223..003fb01 100644
>>> --- a/meta/conf/bitbake.conf
>>> +++ b/meta/conf/bitbake.conf
>>> @@ -36,6 +36,7 @@ S = "${WORKDIR}/${P}"
>>>   AUTOREV = "${@bb.fetch2.get_autorev(d)}"
>>>   SRC_URI = "file://${FILE}"
>>>   STAMP = "${TMPDIR}/stamps/${PF}"
>>> +STAMPCLEAN = "${STAMP}*"
>>
>> I would rather go with "${TMPDIR}/stamps/${PN}-*".
>>
>> OE/YP remove the stamps of other recipe versions as well.
>>
>> This is what OE/YP does:
>>
>>    STAMPS_DIR ?= "${TMPDIR}/stamps"
>>    STAMP = 
>> "${STAMPS_DIR}/${MULTIMACH_TARGET_SYS}/${PN}/${EXTENDPE}${PV}-${PR}"
>>    STAMPCLEAN = "${STAMPS_DIR}/${MULTIMACH_TARGET_SYS}/${PN}/*-*"
> 
> What would be the scenario where this helps? Just trying to understand.

Good question, that I don't have an answer for. I looked at the commit 
messages, but it seems to be that this part of the STAMPCLEAN variable 
goes back to the first commit that introduced it.

It is probably just a general philosophy, should different versions of 
the same recipe invalidate the stamps of each other or not.

regards,
Claudius

-- 
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-54 Fax: (+49)-8142-66989-80 Email: ch@denx.de