* Consistency and use cases for IMAGE_FSTYPES
@ 2012-03-09 21:39 Tom Rini
2012-03-10 6:29 ` Koen Kooi
2012-03-22 23:26 ` Richard Purdie
0 siblings, 2 replies; 12+ messages in thread
From: Tom Rini @ 2012-03-09 21:39 UTC (permalink / raw)
To: Patches and discussions about the oe-core layer
Hey all,
Over in meta-ti I kicked off a discussion
(https://lists.yoctoproject.org/pipermail/meta-ti/2012-March/000779.html)
about if we should be using '?=' or '+=' with IMAGE_FSTYPES in the
machine conf files. This has been discussed a little bit before
(http://thread.gmane.org/gmane.comp.handhelds.openembedded.core/2060/focus=2061).
The problem is we have the following and I believe ultimately
conflicting use cases:
- The machine needs to say 'I need or support the following formats'
- The distro needs to say 'I always want format X'
- The user needs to say 'I know best, give me only format X'
- The user needs to say 'I know best, give me what you support + X'
We achieve this today, but not very nicely. For the last one you need
to start playing order of operations games and that's just not nice.
My suggestion is that we need to change what the machine.conf sets to
another variable (SOMETHING_IMAGE_FSTYPES), and a default
IMAGE_FSTYPES becomes ?= ${SOMETHING_IMAGE_FSTYPES}. Distros, and
users can then work more easily with their use cases. Comments?
--
Tom
^ permalink raw reply [flat|nested] 12+ messages in thread* Re: Consistency and use cases for IMAGE_FSTYPES 2012-03-09 21:39 Consistency and use cases for IMAGE_FSTYPES Tom Rini @ 2012-03-10 6:29 ` Koen Kooi 2012-03-12 20:25 ` Denys Dmytriyenko 2012-03-22 23:26 ` Richard Purdie 1 sibling, 1 reply; 12+ messages in thread From: Koen Kooi @ 2012-03-10 6:29 UTC (permalink / raw) To: Patches and discussions about the oe-core layer Op 9 mrt. 2012, om 22:39 heeft Tom Rini het volgende geschreven: > Hey all, > > Over in meta-ti I kicked off a discussion > (https://lists.yoctoproject.org/pipermail/meta-ti/2012-March/000779.html) > about if we should be using '?=' or '+=' with IMAGE_FSTYPES in the > machine conf files. This has been discussed a little bit before > (http://thread.gmane.org/gmane.comp.handhelds.openembedded.core/2060/focus=2061). > The problem is we have the following and I believe ultimately > conflicting use cases: > - The machine needs to say 'I need or support the following formats' > - The distro needs to say 'I always want format X' > - The user needs to say 'I know best, give me only format X' > - The user needs to say 'I know best, give me what you support + X' > > We achieve this today, but not very nicely. For the last one you need > to start playing order of operations games and that's just not nice. > My suggestion is that we need to change what the machine.conf sets to > another variable (SOMETHING_IMAGE_FSTYPES), and a default > IMAGE_FSTYPES becomes ?= ${SOMETHING_IMAGE_FSTYPES}. Distros, and > users can then work more easily with their use cases. Comments? EXTRA_IMAGE_FSTYPES! regards, Koen ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Consistency and use cases for IMAGE_FSTYPES 2012-03-10 6:29 ` Koen Kooi @ 2012-03-12 20:25 ` Denys Dmytriyenko 2012-03-22 21:40 ` Tom Rini 0 siblings, 1 reply; 12+ messages in thread From: Denys Dmytriyenko @ 2012-03-12 20:25 UTC (permalink / raw) To: Patches and discussions about the oe-core layer On Sat, Mar 10, 2012 at 07:29:20AM +0100, Koen Kooi wrote: > > Op 9 mrt. 2012, om 22:39 heeft Tom Rini het volgende geschreven: > > > Hey all, > > > > Over in meta-ti I kicked off a discussion > > (https://lists.yoctoproject.org/pipermail/meta-ti/2012-March/000779.html) > > about if we should be using '?=' or '+=' with IMAGE_FSTYPES in the > > machine conf files. This has been discussed a little bit before > > (http://thread.gmane.org/gmane.comp.handhelds.openembedded.core/2060/focus=2061). > > The problem is we have the following and I believe ultimately > > conflicting use cases: > > - The machine needs to say 'I need or support the following formats' > > - The distro needs to say 'I always want format X' > > - The user needs to say 'I know best, give me only format X' > > - The user needs to say 'I know best, give me what you support + X' > > > > We achieve this today, but not very nicely. For the last one you need > > to start playing order of operations games and that's just not nice. > > My suggestion is that we need to change what the machine.conf sets to > > another variable (SOMETHING_IMAGE_FSTYPES), and a default > > IMAGE_FSTYPES becomes ?= ${SOMETHING_IMAGE_FSTYPES}. Distros, and > > users can then work more easily with their use cases. Comments? > > EXTRA_IMAGE_FSTYPES! +1 MACHINE_IMAGE_FSTYPES + DISTRO_IMAGE_FSTYPES? jk :) -- Denys ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Consistency and use cases for IMAGE_FSTYPES 2012-03-12 20:25 ` Denys Dmytriyenko @ 2012-03-22 21:40 ` Tom Rini 0 siblings, 0 replies; 12+ messages in thread From: Tom Rini @ 2012-03-22 21:40 UTC (permalink / raw) To: Patches and discussions about the oe-core layer On Mon, Mar 12, 2012 at 04:25:24PM -0400, Denys Dmytriyenko wrote: > On Sat, Mar 10, 2012 at 07:29:20AM +0100, Koen Kooi wrote: > > > > Op 9 mrt. 2012, om 22:39 heeft Tom Rini het volgende geschreven: > > > > > Hey all, > > > > > > Over in meta-ti I kicked off a discussion > > > (https://lists.yoctoproject.org/pipermail/meta-ti/2012-March/000779.html) > > > about if we should be using '?=' or '+=' with IMAGE_FSTYPES in the > > > machine conf files. This has been discussed a little bit before > > > (http://thread.gmane.org/gmane.comp.handhelds.openembedded.core/2060/focus=2061). > > > The problem is we have the following and I believe ultimately > > > conflicting use cases: > > > - The machine needs to say 'I need or support the following formats' > > > - The distro needs to say 'I always want format X' > > > - The user needs to say 'I know best, give me only format X' > > > - The user needs to say 'I know best, give me what you support + X' > > > > > > We achieve this today, but not very nicely. For the last one you need > > > to start playing order of operations games and that's just not nice. > > > My suggestion is that we need to change what the machine.conf sets to > > > another variable (SOMETHING_IMAGE_FSTYPES), and a default > > > IMAGE_FSTYPES becomes ?= ${SOMETHING_IMAGE_FSTYPES}. Distros, and > > > users can then work more easily with their use cases. Comments? > > > > EXTRA_IMAGE_FSTYPES! > > +1 > > MACHINE_IMAGE_FSTYPES + DISTRO_IMAGE_FSTYPES? jk :) Having gotten no other feedback (*cough*) I'll post a patch for SOMETHING_IMAGE_FSTYPES (I'm leaning MACHINE_IMAGE_FSTYPES, sorry Koen :)) sometime tomorrow I think. -- Tom ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Consistency and use cases for IMAGE_FSTYPES 2012-03-09 21:39 Consistency and use cases for IMAGE_FSTYPES Tom Rini 2012-03-10 6:29 ` Koen Kooi @ 2012-03-22 23:26 ` Richard Purdie 2012-03-22 23:53 ` Denys Dmytriyenko 1 sibling, 1 reply; 12+ messages in thread From: Richard Purdie @ 2012-03-22 23:26 UTC (permalink / raw) To: Patches and discussions about the oe-core layer On Fri, 2012-03-09 at 14:39 -0700, Tom Rini wrote: > Hey all, > > Over in meta-ti I kicked off a discussion > (https://lists.yoctoproject.org/pipermail/meta-ti/2012-March/000779.html) > about if we should be using '?=' or '+=' with IMAGE_FSTYPES in the > machine conf files. This has been discussed a little bit before > (http://thread.gmane.org/gmane.comp.handhelds.openembedded.core/2060/focus=2061). > The problem is we have the following and I believe ultimately > conflicting use cases: I've been under the impression that we decided upon: > - The machine needs to say 'I need or support the following formats' so the machine starts and sets: IMAGE_FSTYPES = "xxxx" > - The distro needs to say 'I always want format X' so the distro can do: IMAGE_FSTYPES += " yyy" > - The user needs to say 'I know best, give me only format X' So the user can do: IMAGE_FSTYPES = "X" > - The user needs to say 'I know best, give me what you support + X' IMAGE_FSTYPES += " X" > We achieve this today, but not very nicely. For the last one you need > to start playing order of operations games and that's just not nice. > My suggestion is that we need to change what the machine.conf sets to > another variable (SOMETHING_IMAGE_FSTYPES), and a default > IMAGE_FSTYPES becomes ?= ${SOMETHING_IMAGE_FSTYPES}. Distros, and > users can then work more easily with their use cases. Comments? I think this is overcomplicating things. Yes the order is important but lets just document the above? Cheers, Richard ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Consistency and use cases for IMAGE_FSTYPES 2012-03-22 23:26 ` Richard Purdie @ 2012-03-22 23:53 ` Denys Dmytriyenko 2012-03-23 1:14 ` Richard Purdie 0 siblings, 1 reply; 12+ messages in thread From: Denys Dmytriyenko @ 2012-03-22 23:53 UTC (permalink / raw) To: Patches and discussions about the oe-core layer On Thu, Mar 22, 2012 at 11:26:24PM +0000, Richard Purdie wrote: > On Fri, 2012-03-09 at 14:39 -0700, Tom Rini wrote: > > Hey all, > > > > Over in meta-ti I kicked off a discussion > > (https://lists.yoctoproject.org/pipermail/meta-ti/2012-March/000779.html) > > about if we should be using '?=' or '+=' with IMAGE_FSTYPES in the > > machine conf files. This has been discussed a little bit before > > (http://thread.gmane.org/gmane.comp.handhelds.openembedded.core/2060/focus=2061). > > The problem is we have the following and I believe ultimately > > conflicting use cases: > > I've been under the impression that we decided upon: > > > - The machine needs to say 'I need or support the following formats' > > so the machine starts and sets: > > IMAGE_FSTYPES = "xxxx" > > > - The distro needs to say 'I always want format X' > > so the distro can do: > > IMAGE_FSTYPES += " yyy" > > > - The user needs to say 'I know best, give me only format X' > > So the user can do: > > IMAGE_FSTYPES = "X" Richard, Since local.conf gets parsed before machine.conf and distro.conf, the user needs to do this override: IMAGE_FSTYPES_local = "X" Otherwise machine.conf will always overwrite it with "xxxx" with its unconditional assignment. -- Denys > > - The user needs to say 'I know best, give me what you support + X' > > IMAGE_FSTYPES += " X" > > > We achieve this today, but not very nicely. For the last one you need > > to start playing order of operations games and that's just not nice. > > My suggestion is that we need to change what the machine.conf sets to > > another variable (SOMETHING_IMAGE_FSTYPES), and a default > > IMAGE_FSTYPES becomes ?= ${SOMETHING_IMAGE_FSTYPES}. Distros, and > > users can then work more easily with their use cases. Comments? > > I think this is overcomplicating things. Yes the order is important but > lets just document the above? ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Consistency and use cases for IMAGE_FSTYPES 2012-03-22 23:53 ` Denys Dmytriyenko @ 2012-03-23 1:14 ` Richard Purdie 2012-03-23 15:48 ` Tom Rini 0 siblings, 1 reply; 12+ messages in thread From: Richard Purdie @ 2012-03-23 1:14 UTC (permalink / raw) To: Patches and discussions about the oe-core layer On Thu, 2012-03-22 at 19:53 -0400, Denys Dmytriyenko wrote: > On Thu, Mar 22, 2012 at 11:26:24PM +0000, Richard Purdie wrote: > > On Fri, 2012-03-09 at 14:39 -0700, Tom Rini wrote: > > > Hey all, > > > > > > Over in meta-ti I kicked off a discussion > > > (https://lists.yoctoproject.org/pipermail/meta-ti/2012-March/000779.html) > > > about if we should be using '?=' or '+=' with IMAGE_FSTYPES in the > > > machine conf files. This has been discussed a little bit before > > > (http://thread.gmane.org/gmane.comp.handhelds.openembedded.core/2060/focus=2061). > > > The problem is we have the following and I believe ultimately > > > conflicting use cases: > > > > I've been under the impression that we decided upon: > > > > > - The machine needs to say 'I need or support the following formats' > > > > so the machine starts and sets: > > > > IMAGE_FSTYPES = "xxxx" > > > > > - The distro needs to say 'I always want format X' > > > > so the distro can do: > > > > IMAGE_FSTYPES += " yyy" > > > > > - The user needs to say 'I know best, give me only format X' > > > > So the user can do: > > > > IMAGE_FSTYPES = "X" > > Since local.conf gets parsed before machine.conf and distro.conf, the user > needs to do this override: > > IMAGE_FSTYPES_local = "X" > > Otherwise machine.conf will always overwrite it with "xxxx" with its > unconditional assignment. Right, I'd forgotten that little detail :/. It actually makes me wonder if our include order is the right one but now isn't the time to try changing that. I agree the neatest way to change it is probably something like MACHINE_FSTYPES. I do worry a lot about backwards compatibility though and I'd also point out where we're at in the release cycle (bug fix only). Cheers, Richard ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Consistency and use cases for IMAGE_FSTYPES 2012-03-23 1:14 ` Richard Purdie @ 2012-03-23 15:48 ` Tom Rini 2012-03-23 16:17 ` Darren Hart 0 siblings, 1 reply; 12+ messages in thread From: Tom Rini @ 2012-03-23 15:48 UTC (permalink / raw) To: Patches and discussions about the oe-core layer [-- Attachment #1: Type: text/plain, Size: 2186 bytes --] On Fri, Mar 23, 2012 at 01:14:24AM +0000, Richard Purdie wrote: > On Thu, 2012-03-22 at 19:53 -0400, Denys Dmytriyenko wrote: > > On Thu, Mar 22, 2012 at 11:26:24PM +0000, Richard Purdie wrote: > > > On Fri, 2012-03-09 at 14:39 -0700, Tom Rini wrote: > > > > Hey all, > > > > > > > > Over in meta-ti I kicked off a discussion > > > > (https://lists.yoctoproject.org/pipermail/meta-ti/2012-March/000779.html) > > > > about if we should be using '?=' or '+=' with IMAGE_FSTYPES in the > > > > machine conf files. This has been discussed a little bit before > > > > (http://thread.gmane.org/gmane.comp.handhelds.openembedded.core/2060/focus=2061). > > > > The problem is we have the following and I believe ultimately > > > > conflicting use cases: > > > > > > I've been under the impression that we decided upon: > > > > > > > - The machine needs to say 'I need or support the following formats' > > > > > > so the machine starts and sets: > > > > > > IMAGE_FSTYPES = "xxxx" > > > > > > > - The distro needs to say 'I always want format X' > > > > > > so the distro can do: > > > > > > IMAGE_FSTYPES += " yyy" > > > > > > > - The user needs to say 'I know best, give me only format X' > > > > > > So the user can do: > > > > > > IMAGE_FSTYPES = "X" > > > > Since local.conf gets parsed before machine.conf and distro.conf, the user > > needs to do this override: > > > > IMAGE_FSTYPES_local = "X" > > > > Otherwise machine.conf will always overwrite it with "xxxx" with its > > unconditional assignment. > > Right, I'd forgotten that little detail :/. > > It actually makes me wonder if our include order is the right one but > now isn't the time to try changing that. > > I agree the neatest way to change it is probably something like > MACHINE_FSTYPES. I do worry a lot about backwards compatibility though > and I'd also point out where we're at in the release cycle (bug fix > only). Well, one problem that would make this a bugfix is that no one does what you say we agreed on today. oe-core has qemu.inc using ?=, meta-intel is using += and meta-ti is mixed (which is what got this started). -- Tom [-- Attachment #2: Digital signature --] [-- Type: application/pgp-signature, Size: 205 bytes --] ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Consistency and use cases for IMAGE_FSTYPES 2012-03-23 15:48 ` Tom Rini @ 2012-03-23 16:17 ` Darren Hart 2012-03-23 16:29 ` Tom Rini 0 siblings, 1 reply; 12+ messages in thread From: Darren Hart @ 2012-03-23 16:17 UTC (permalink / raw) To: Patches and discussions about the oe-core layer On 03/23/2012 08:48 AM, Tom Rini wrote: > On Fri, Mar 23, 2012 at 01:14:24AM +0000, Richard Purdie wrote: >> On Thu, 2012-03-22 at 19:53 -0400, Denys Dmytriyenko wrote: >>> On Thu, Mar 22, 2012 at 11:26:24PM +0000, Richard Purdie wrote: >>>> On Fri, 2012-03-09 at 14:39 -0700, Tom Rini wrote: >>>>> Hey all, >>>>> >>>>> Over in meta-ti I kicked off a discussion >>>>> (https://lists.yoctoproject.org/pipermail/meta-ti/2012-March/000779.html) >>>>> about if we should be using '?=' or '+=' with IMAGE_FSTYPES in the >>>>> machine conf files. This has been discussed a little bit before >>>>> (http://thread.gmane.org/gmane.comp.handhelds.openembedded.core/2060/focus=2061). >>>>> The problem is we have the following and I believe ultimately >>>>> conflicting use cases: >>>> >>>> I've been under the impression that we decided upon: >>>> >>>>> - The machine needs to say 'I need or support the following formats' >>>> >>>> so the machine starts and sets: >>>> >>>> IMAGE_FSTYPES = "xxxx" >>>> >>>>> - The distro needs to say 'I always want format X' >>>> >>>> so the distro can do: >>>> >>>> IMAGE_FSTYPES += " yyy" >>>> >>>>> - The user needs to say 'I know best, give me only format X' >>>> >>>> So the user can do: >>>> >>>> IMAGE_FSTYPES = "X" >>> >>> Since local.conf gets parsed before machine.conf and distro.conf, the user >>> needs to do this override: >>> >>> IMAGE_FSTYPES_local = "X" >>> >>> Otherwise machine.conf will always overwrite it with "xxxx" with its >>> unconditional assignment. >> >> Right, I'd forgotten that little detail :/. >> >> It actually makes me wonder if our include order is the right one but >> now isn't the time to try changing that. >> >> I agree the neatest way to change it is probably something like >> MACHINE_FSTYPES. I do worry a lot about backwards compatibility though >> and I'd also point out where we're at in the release cycle (bug fix >> only). > > Well, one problem that would make this a bugfix is that no one does what > you say we agreed on today. oe-core has qemu.inc using ?=, meta-intel > is using += and meta-ti is mixed (which is what got this started). > > Is this causing any nasty failures right now, or is it in the "this is a confusing mess and it would be nice to get it cleaned up" bucket? If the latter, I think I'd prefer to wait a bit an clean up the local.conf/machine.conf IMAGE_FSTYPES clobbering issue. If this isn't really fixable (for whatever requirements bitbake has on load/parse order of config files), then Koen's EXTRA_IMAGE_FSTYPES seems like the most consistent mechanism with other things, like CORE_IMAGE_EXTRA_INSTALL (OK, maybe IMAGE_EXTRA_FSTYPES ?). So the default becomes: IMAGE_FSTYPES ?= ${IMAGE_EXTRA_FSTYPES} and DISTROs might define that as: IMAGE_FSTYPES += "yyy" and users can update local.conf to be: IMAGE_FSTYPES = "X" But, doesn't this meant the DISTRO append will still change the IMAGE_FSTYPES to "X yyy" even though the user intended "only X"? -- Darren Hart Intel Open Source Technology Center Yocto Project - Linux Kernel ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Consistency and use cases for IMAGE_FSTYPES 2012-03-23 16:17 ` Darren Hart @ 2012-03-23 16:29 ` Tom Rini 2012-03-23 16:34 ` Darren Hart 0 siblings, 1 reply; 12+ messages in thread From: Tom Rini @ 2012-03-23 16:29 UTC (permalink / raw) To: Darren Hart; +Cc: Patches and discussions about the oe-core layer [-- Attachment #1: Type: text/plain, Size: 3858 bytes --] On Fri, Mar 23, 2012 at 09:17:20AM -0700, Darren Hart wrote: > On 03/23/2012 08:48 AM, Tom Rini wrote: > > On Fri, Mar 23, 2012 at 01:14:24AM +0000, Richard Purdie wrote: > >> On Thu, 2012-03-22 at 19:53 -0400, Denys Dmytriyenko wrote: > >>> On Thu, Mar 22, 2012 at 11:26:24PM +0000, Richard Purdie wrote: > >>>> On Fri, 2012-03-09 at 14:39 -0700, Tom Rini wrote: > >>>>> Hey all, > >>>>> > >>>>> Over in meta-ti I kicked off a discussion > >>>>> (https://lists.yoctoproject.org/pipermail/meta-ti/2012-March/000779.html) > >>>>> about if we should be using '?=' or '+=' with IMAGE_FSTYPES in the > >>>>> machine conf files. This has been discussed a little bit before > >>>>> (http://thread.gmane.org/gmane.comp.handhelds.openembedded.core/2060/focus=2061). > >>>>> The problem is we have the following and I believe ultimately > >>>>> conflicting use cases: > >>>> > >>>> I've been under the impression that we decided upon: > >>>> > >>>>> - The machine needs to say 'I need or support the following formats' > >>>> > >>>> so the machine starts and sets: > >>>> > >>>> IMAGE_FSTYPES = "xxxx" > >>>> > >>>>> - The distro needs to say 'I always want format X' > >>>> > >>>> so the distro can do: > >>>> > >>>> IMAGE_FSTYPES += " yyy" > >>>> > >>>>> - The user needs to say 'I know best, give me only format X' > >>>> > >>>> So the user can do: > >>>> > >>>> IMAGE_FSTYPES = "X" > >>> > >>> Since local.conf gets parsed before machine.conf and distro.conf, the user > >>> needs to do this override: > >>> > >>> IMAGE_FSTYPES_local = "X" > >>> > >>> Otherwise machine.conf will always overwrite it with "xxxx" with its > >>> unconditional assignment. > >> > >> Right, I'd forgotten that little detail :/. > >> > >> It actually makes me wonder if our include order is the right one but > >> now isn't the time to try changing that. > >> > >> I agree the neatest way to change it is probably something like > >> MACHINE_FSTYPES. I do worry a lot about backwards compatibility though > >> and I'd also point out where we're at in the release cycle (bug fix > >> only). > > > > Well, one problem that would make this a bugfix is that no one does what > > you say we agreed on today. oe-core has qemu.inc using ?=, meta-intel > > is using += and meta-ti is mixed (which is what got this started). > > > > > > Is this causing any nasty failures right now, or is it in the "this is a > confusing mess and it would be nice to get it cleaned up" bucket? If the > latter, I think I'd prefer to wait a bit an clean up the > local.conf/machine.conf IMAGE_FSTYPES clobbering issue. Well, I found this as part of adding UBI support for a board and it wasn't sticking. I'd go so far as to say that for a release, we really need to pick a standard, document and follow it. If it's machine.conf does =, everyone else does += and user's have to do _local =, fine, it sucks but it's documented and consistent on all of the BSP layers. > If this isn't really fixable (for whatever requirements bitbake has on > load/parse order of config files), then Koen's EXTRA_IMAGE_FSTYPES seems > like the most consistent mechanism with other things, like > CORE_IMAGE_EXTRA_INSTALL (OK, maybe IMAGE_EXTRA_FSTYPES ?). > > So the default becomes: > > IMAGE_FSTYPES ?= ${IMAGE_EXTRA_FSTYPES} > > and DISTROs might define that as: > > IMAGE_FSTYPES += "yyy" > > and users can update local.conf to be: > > IMAGE_FSTYPES = "X" > > But, doesn't this meant the DISTRO append will still change the > IMAGE_FSTYPES to "X yyy" even though the user intended "only X"? How about: bitbake.conf: IMAGE_FSTYPES ??= ${IMAGE_EXTRA_FSTYPES} distro.conf: IMAGE_FSTYPES ?= "yyy ${IMAGE_EXTRA_FSTYPES}" local.conf: IMAGE_FSTYPES = "X" Or am I forgetting the magic of ??= again... -- Tom [-- Attachment #2: Digital signature --] [-- Type: application/pgp-signature, Size: 205 bytes --] ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Consistency and use cases for IMAGE_FSTYPES 2012-03-23 16:29 ` Tom Rini @ 2012-03-23 16:34 ` Darren Hart 2012-03-23 17:03 ` Tom Rini 0 siblings, 1 reply; 12+ messages in thread From: Darren Hart @ 2012-03-23 16:34 UTC (permalink / raw) To: Tom Rini; +Cc: Patches and discussions about the oe-core layer -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 03/23/2012 09:29 AM, Tom Rini wrote: > On Fri, Mar 23, 2012 at 09:17:20AM -0700, Darren Hart wrote: >> On 03/23/2012 08:48 AM, Tom Rini wrote: >>> On Fri, Mar 23, 2012 at 01:14:24AM +0000, Richard Purdie >>> wrote: >>>> On Thu, 2012-03-22 at 19:53 -0400, Denys Dmytriyenko wrote: >>>>> On Thu, Mar 22, 2012 at 11:26:24PM +0000, Richard Purdie >>>>> wrote: >>>>>> On Fri, 2012-03-09 at 14:39 -0700, Tom Rini wrote: >>>>>>> Hey all, >>>>>>> >>>>>>> Over in meta-ti I kicked off a discussion >>>>>>> (https://lists.yoctoproject.org/pipermail/meta-ti/2012-March/000779.html) >>>>>>> >>>>>>> about if we should be using '?=' or '+=' with IMAGE_FSTYPES in the >>>>>>> machine conf files. This has been discussed a little >>>>>>> bit before >>>>>>> (http://thread.gmane.org/gmane.comp.handhelds.openembedded.core/2060/focus=2061). >>>>>>> >>>>>>> The problem is we have the following and I believe ultimately >>>>>>> conflicting use cases: >>>>>> >>>>>> I've been under the impression that we decided upon: >>>>>> >>>>>>> - The machine needs to say 'I need or support the >>>>>>> following formats' >>>>>> >>>>>> so the machine starts and sets: >>>>>> >>>>>> IMAGE_FSTYPES = "xxxx" >>>>>> >>>>>>> - The distro needs to say 'I always want format X' >>>>>> >>>>>> so the distro can do: >>>>>> >>>>>> IMAGE_FSTYPES += " yyy" >>>>>> >>>>>>> - The user needs to say 'I know best, give me only >>>>>>> format X' >>>>>> >>>>>> So the user can do: >>>>>> >>>>>> IMAGE_FSTYPES = "X" >>>>> >>>>> Since local.conf gets parsed before machine.conf and >>>>> distro.conf, the user needs to do this override: >>>>> >>>>> IMAGE_FSTYPES_local = "X" >>>>> >>>>> Otherwise machine.conf will always overwrite it with "xxxx" >>>>> with its unconditional assignment. >>>> >>>> Right, I'd forgotten that little detail :/. >>>> >>>> It actually makes me wonder if our include order is the right >>>> one but now isn't the time to try changing that. >>>> >>>> I agree the neatest way to change it is probably something >>>> like MACHINE_FSTYPES. I do worry a lot about backwards >>>> compatibility though and I'd also point out where we're at in >>>> the release cycle (bug fix only). >>> >>> Well, one problem that would make this a bugfix is that no one >>> does what you say we agreed on today. oe-core has qemu.inc >>> using ?=, meta-intel is using += and meta-ti is mixed (which is >>> what got this started). >>> >>> >> >> Is this causing any nasty failures right now, or is it in the >> "this is a confusing mess and it would be nice to get it cleaned >> up" bucket? If the latter, I think I'd prefer to wait a bit an >> clean up the local.conf/machine.conf IMAGE_FSTYPES clobbering >> issue. > > Well, I found this as part of adding UBI support for a board and > it wasn't sticking. > > I'd go so far as to say that for a release, we really need to pick > a standard, document and follow it. If it's machine.conf does =, > everyone else does += and user's have to do _local =, fine, it > sucks but it's documented and consistent on all of the BSP layers. > >> If this isn't really fixable (for whatever requirements bitbake >> has on load/parse order of config files), then Koen's >> EXTRA_IMAGE_FSTYPES seems like the most consistent mechanism with >> other things, like CORE_IMAGE_EXTRA_INSTALL (OK, maybe >> IMAGE_EXTRA_FSTYPES ?). >> >> So the default becomes: >> >> IMAGE_FSTYPES ?= ${IMAGE_EXTRA_FSTYPES} >> >> and DISTROs might define that as: >> >> IMAGE_FSTYPES += "yyy" >> >> and users can update local.conf to be: >> >> IMAGE_FSTYPES = "X" >> >> But, doesn't this meant the DISTRO append will still change the >> IMAGE_FSTYPES to "X yyy" even though the user intended "only X"? > > How about: bitbake.conf: IMAGE_FSTYPES ??= ${IMAGE_EXTRA_FSTYPES} > distro.conf: IMAGE_FSTYPES ?= "yyy ${IMAGE_EXTRA_FSTYPES}" > local.conf: IMAGE_FSTYPES = "X" > > Or am I forgetting the magic of ??= again... > What would machine.conf do in this scenario? IMAGE_FSTYPES_append_machine = "Z" ? or IMAGE_EXTRA_FSTYPES = "Z" - -- Darren Hart Intel Open Source Technology Center Yocto Project - Linux Kernel -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iQEcBAEBAgAGBQJPbKXvAAoJEKbMaAwKp364kN8IAKygG06LhMH7FI6E4L0xuX/U i6szfRYOX2mWz6OAf/pPJtyvDjtwpgpLvRQEpOfIIVPciOlw/aTCZW9kGRG6hQtl UxnNfyp25tqc7IFTm5AXVRvfMGlI+V7tXoPK117i/EXyuRQbPood5QUN6YUR+NXF Bw0gxhYr/JSZtKiBCIRRbO7o5SSGpcDydOkG7RIsA3XS2dT5YzuGfFWTKbYyAct/ g97aqvxXxL3h+/kr5rJ9dR+qYjCsmkyqT6ePDzXlkzQEoQUHs+7u+u0Au8Sg+Sxg W53G3MoOlzSjLojhTlTYfzuwxfEv7Jkg0q6hYOGUSlGPQ5Fuc8UfAiSi5MAQVL0= =92cW -----END PGP SIGNATURE----- ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: Consistency and use cases for IMAGE_FSTYPES 2012-03-23 16:34 ` Darren Hart @ 2012-03-23 17:03 ` Tom Rini 0 siblings, 0 replies; 12+ messages in thread From: Tom Rini @ 2012-03-23 17:03 UTC (permalink / raw) To: Darren Hart; +Cc: Patches and discussions about the oe-core layer [-- Attachment #1: Type: text/plain, Size: 5059 bytes --] On Fri, Mar 23, 2012 at 09:34:02AM -0700, Darren Hart wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > > > On 03/23/2012 09:29 AM, Tom Rini wrote: > > On Fri, Mar 23, 2012 at 09:17:20AM -0700, Darren Hart wrote: > >> On 03/23/2012 08:48 AM, Tom Rini wrote: > >>> On Fri, Mar 23, 2012 at 01:14:24AM +0000, Richard Purdie > >>> wrote: > >>>> On Thu, 2012-03-22 at 19:53 -0400, Denys Dmytriyenko wrote: > >>>>> On Thu, Mar 22, 2012 at 11:26:24PM +0000, Richard Purdie > >>>>> wrote: > >>>>>> On Fri, 2012-03-09 at 14:39 -0700, Tom Rini wrote: > >>>>>>> Hey all, > >>>>>>> > >>>>>>> Over in meta-ti I kicked off a discussion > >>>>>>> (https://lists.yoctoproject.org/pipermail/meta-ti/2012-March/000779.html) > >>>>>>> > >>>>>>> > about if we should be using '?=' or '+=' with IMAGE_FSTYPES in the > >>>>>>> machine conf files. This has been discussed a little > >>>>>>> bit before > >>>>>>> (http://thread.gmane.org/gmane.comp.handhelds.openembedded.core/2060/focus=2061). > >>>>>>> > >>>>>>> > The problem is we have the following and I believe ultimately > >>>>>>> conflicting use cases: > >>>>>> > >>>>>> I've been under the impression that we decided upon: > >>>>>> > >>>>>>> - The machine needs to say 'I need or support the > >>>>>>> following formats' > >>>>>> > >>>>>> so the machine starts and sets: > >>>>>> > >>>>>> IMAGE_FSTYPES = "xxxx" > >>>>>> > >>>>>>> - The distro needs to say 'I always want format X' > >>>>>> > >>>>>> so the distro can do: > >>>>>> > >>>>>> IMAGE_FSTYPES += " yyy" > >>>>>> > >>>>>>> - The user needs to say 'I know best, give me only > >>>>>>> format X' > >>>>>> > >>>>>> So the user can do: > >>>>>> > >>>>>> IMAGE_FSTYPES = "X" > >>>>> > >>>>> Since local.conf gets parsed before machine.conf and > >>>>> distro.conf, the user needs to do this override: > >>>>> > >>>>> IMAGE_FSTYPES_local = "X" > >>>>> > >>>>> Otherwise machine.conf will always overwrite it with "xxxx" > >>>>> with its unconditional assignment. > >>>> > >>>> Right, I'd forgotten that little detail :/. > >>>> > >>>> It actually makes me wonder if our include order is the right > >>>> one but now isn't the time to try changing that. > >>>> > >>>> I agree the neatest way to change it is probably something > >>>> like MACHINE_FSTYPES. I do worry a lot about backwards > >>>> compatibility though and I'd also point out where we're at in > >>>> the release cycle (bug fix only). > >>> > >>> Well, one problem that would make this a bugfix is that no one > >>> does what you say we agreed on today. oe-core has qemu.inc > >>> using ?=, meta-intel is using += and meta-ti is mixed (which is > >>> what got this started). > >>> > >>> > >> > >> Is this causing any nasty failures right now, or is it in the > >> "this is a confusing mess and it would be nice to get it cleaned > >> up" bucket? If the latter, I think I'd prefer to wait a bit an > >> clean up the local.conf/machine.conf IMAGE_FSTYPES clobbering > >> issue. > > > > Well, I found this as part of adding UBI support for a board and > > it wasn't sticking. > > > > I'd go so far as to say that for a release, we really need to pick > > a standard, document and follow it. If it's machine.conf does =, > > everyone else does += and user's have to do _local =, fine, it > > sucks but it's documented and consistent on all of the BSP layers. > > > >> If this isn't really fixable (for whatever requirements bitbake > >> has on load/parse order of config files), then Koen's > >> EXTRA_IMAGE_FSTYPES seems like the most consistent mechanism with > >> other things, like CORE_IMAGE_EXTRA_INSTALL (OK, maybe > >> IMAGE_EXTRA_FSTYPES ?). > >> > >> So the default becomes: > >> > >> IMAGE_FSTYPES ?= ${IMAGE_EXTRA_FSTYPES} > >> > >> and DISTROs might define that as: > >> > >> IMAGE_FSTYPES += "yyy" > >> > >> and users can update local.conf to be: > >> > >> IMAGE_FSTYPES = "X" > >> > >> But, doesn't this meant the DISTRO append will still change the > >> IMAGE_FSTYPES to "X yyy" even though the user intended "only X"? > > > > How about: bitbake.conf: IMAGE_FSTYPES ??= ${IMAGE_EXTRA_FSTYPES} > > distro.conf: IMAGE_FSTYPES ?= "yyy ${IMAGE_EXTRA_FSTYPES}" > > local.conf: IMAGE_FSTYPES = "X" > > > > Or am I forgetting the magic of ??= again... > > > > What would machine.conf do in this scenario? OK, lets test things out. bitbake.conf, distro.conf set as above (with machine.conf providing IMAGE_EXTRA_FSTYPES). > IMAGE_FSTYPES_append_machine = "Z" ? With this in local.conf, IMAGE_FSTYPES = "distro machineZ" (the fun of _append semantics, you would have wanted " Z"). > IMAGE_EXTRA_FSTYPES = "Z" Again in local.conf, this is ignored and it's the original problem. Making this IMAGE_EXTRA_FSTYPES_local = "Z" is also ignored because local overrides were removed in 83ce96f (really? ok..). Making this IMAGE_EXTRA_FSTYPES_forcevariable = "Z" yields IMAGE_FSTYPES = "distro Z". -- Tom [-- Attachment #2: Digital signature --] [-- Type: application/pgp-signature, Size: 205 bytes --] ^ permalink raw reply [flat|nested] 12+ messages in thread
end of thread, other threads:[~2012-03-23 17:12 UTC | newest] Thread overview: 12+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2012-03-09 21:39 Consistency and use cases for IMAGE_FSTYPES Tom Rini 2012-03-10 6:29 ` Koen Kooi 2012-03-12 20:25 ` Denys Dmytriyenko 2012-03-22 21:40 ` Tom Rini 2012-03-22 23:26 ` Richard Purdie 2012-03-22 23:53 ` Denys Dmytriyenko 2012-03-23 1:14 ` Richard Purdie 2012-03-23 15:48 ` Tom Rini 2012-03-23 16:17 ` Darren Hart 2012-03-23 16:29 ` Tom Rini 2012-03-23 16:34 ` Darren Hart 2012-03-23 17:03 ` Tom Rini
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.