Message ID | 20221111225827.2412062-1-richard.purdie@linuxfoundation.org |
---|---|
State | Accepted, archived |
Commit | d3eb4531aae28a07cb7e52ed5fe1102445d2effd |
Headers | show |
Series | sanity: Drop data finalize call | expand |
> -----Original Message----- > From: openembedded-core@lists.openembedded.org <openembedded-core@lists.openembedded.org> On Behalf Of Richard Purdie > Sent: den 11 november 2022 23:58 > To: openembedded-core@lists.openembedded.org > Subject: [OE-core] [PATCH] sanity: Drop data finalize call > > This call was effectively like update_data and no longer did anything > in bitbake. Drop it as it is obsolete. > > Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> > --- > meta/classes-global/sanity.bbclass | 11 ++--------- > 1 file changed, 2 insertions(+), 9 deletions(-) > > diff --git a/meta/classes-global/sanity.bbclass b/meta/classes-global/sanity.bbclass > index 8c7d4d08809..6cb0d6cf882 100644 > --- a/meta/classes-global/sanity.bbclass > +++ b/meta/classes-global/sanity.bbclass > @@ -1003,13 +1003,6 @@ def check_sanity(sanity_data): > if status.messages != "": > raise_sanity_error(sanity_data.expand(status.messages), sanity_data, status.network_error) > > -# Create a copy of the datastore and finalise it to ensure appends and > -# overrides are set - the datastore has yet to be finalised at ConfigParsed > -def copy_data(e): > - sanity_data = bb.data.createCopy(e.data) > - sanity_data.finalize() > - return sanity_data > - > addhandler config_reparse_eventhandler > config_reparse_eventhandler[eventmask] = "bb.event.ConfigParsed" > python config_reparse_eventhandler() { > @@ -1020,13 +1013,13 @@ addhandler check_sanity_eventhandler > check_sanity_eventhandler[eventmask] = "bb.event.SanityCheck bb.event.NetworkTest" > python check_sanity_eventhandler() { > if bb.event.getName(e) == "SanityCheck": > - sanity_data = copy_data(e) > + sanity_data = bb.data.createCopy(e.data) For my understanding, wouldn't it be the same to do: sanity_data = bb.data.createCopy(d) I found this in the bitbake documentation, which seems to support my understanding: A global variable "e" is defined, which represents the current event. ... The global datastore is available as "d". In legacy code, you might see "e.data" used to get the datastore. However, realize that "e.data" is deprecated and you should use "d" going forward. Further, is there any reason to use bb.data.createCopy(d) instead of d.createCopy()? AFAICT, the former just calls the latter. > check_sanity(sanity_data) > if e.generateevents: > sanity_data.setVar("SANITY_USE_EVENTS", "1") > bb.event.fire(bb.event.SanityCheckPassed(), e.data) > elif bb.event.getName(e) == "NetworkTest": > - sanity_data = copy_data(e) > + sanity_data = bb.data.createCopy(e.data) > if e.generateevents: > sanity_data.setVar("SANITY_USE_EVENTS", "1") > bb.event.fire(bb.event.NetworkTestFailed() if check_connectivity(sanity_data) else bb.event.NetworkTestPassed(), e.data) > -- > 2.34.1 //Peter
On Tue, 2022-11-15 at 15:10 +0000, Peter Kjellerstedt wrote: > > -----Original Message----- > > From: openembedded-core@lists.openembedded.org > > <openembedded-core@lists.openembedded.org> On Behalf Of Richard > > Purdie > > Sent: den 11 november 2022 23:58 > > To: openembedded-core@lists.openembedded.org > > Subject: [OE-core] [PATCH] sanity: Drop data finalize call > > > > This call was effectively like update_data and no longer did > > anything > > in bitbake. Drop it as it is obsolete. > > > > Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> > > --- > > meta/classes-global/sanity.bbclass | 11 ++--------- > > 1 file changed, 2 insertions(+), 9 deletions(-) > > > > diff --git a/meta/classes-global/sanity.bbclass b/meta/classes- > > global/sanity.bbclass > > index 8c7d4d08809..6cb0d6cf882 100644 > > --- a/meta/classes-global/sanity.bbclass > > +++ b/meta/classes-global/sanity.bbclass > > @@ -1003,13 +1003,6 @@ def check_sanity(sanity_data): > > if status.messages != "": > > raise_sanity_error(sanity_data.expand(status.messages), > > sanity_data, status.network_error) > > > > -# Create a copy of the datastore and finalise it to ensure appends > > and > > -# overrides are set - the datastore has yet to be finalised at > > ConfigParsed > > -def copy_data(e): > > - sanity_data = bb.data.createCopy(e.data) > > - sanity_data.finalize() > > - return sanity_data > > - > > addhandler config_reparse_eventhandler > > config_reparse_eventhandler[eventmask] = "bb.event.ConfigParsed" > > python config_reparse_eventhandler() { > > @@ -1020,13 +1013,13 @@ addhandler check_sanity_eventhandler > > check_sanity_eventhandler[eventmask] = "bb.event.SanityCheck > > bb.event.NetworkTest" > > python check_sanity_eventhandler() { > > if bb.event.getName(e) == "SanityCheck": > > - sanity_data = copy_data(e) > > + sanity_data = bb.data.createCopy(e.data) > > For my understanding, wouldn't it be the same to do: > > sanity_data = bb.data.createCopy(d) Yes. > I found this in the bitbake documentation, which seems to support my > understanding: A global variable "e" is defined, which represents the > current event. ... The global datastore is available as "d". In > legacy > code, you might see "e.data" used to get the datastore. However, > realize > that "e.data" is deprecated and you should use "d" going forward. e.data is how we used to have to access the datastore in event contexts. We did add d to allow standardisation and we do want to remove e.data but haven't cleaned out the old references yet. Arguably I could/should have here but I was already nesting a few issues deep in cleanups so I've left that for another time. > Further, is there any reason to use bb.data.createCopy(d) instead of > d.createCopy()? AFAICT, the former just calls the latter. d.createCopy() is indeed the preferred syntax and functionally the same. Cheers, Richard
diff --git a/meta/classes-global/sanity.bbclass b/meta/classes-global/sanity.bbclass index 8c7d4d08809..6cb0d6cf882 100644 --- a/meta/classes-global/sanity.bbclass +++ b/meta/classes-global/sanity.bbclass @@ -1003,13 +1003,6 @@ def check_sanity(sanity_data): if status.messages != "": raise_sanity_error(sanity_data.expand(status.messages), sanity_data, status.network_error) -# Create a copy of the datastore and finalise it to ensure appends and -# overrides are set - the datastore has yet to be finalised at ConfigParsed -def copy_data(e): - sanity_data = bb.data.createCopy(e.data) - sanity_data.finalize() - return sanity_data - addhandler config_reparse_eventhandler config_reparse_eventhandler[eventmask] = "bb.event.ConfigParsed" python config_reparse_eventhandler() { @@ -1020,13 +1013,13 @@ addhandler check_sanity_eventhandler check_sanity_eventhandler[eventmask] = "bb.event.SanityCheck bb.event.NetworkTest" python check_sanity_eventhandler() { if bb.event.getName(e) == "SanityCheck": - sanity_data = copy_data(e) + sanity_data = bb.data.createCopy(e.data) check_sanity(sanity_data) if e.generateevents: sanity_data.setVar("SANITY_USE_EVENTS", "1") bb.event.fire(bb.event.SanityCheckPassed(), e.data) elif bb.event.getName(e) == "NetworkTest": - sanity_data = copy_data(e) + sanity_data = bb.data.createCopy(e.data) if e.generateevents: sanity_data.setVar("SANITY_USE_EVENTS", "1") bb.event.fire(bb.event.NetworkTestFailed() if check_connectivity(sanity_data) else bb.event.NetworkTestPassed(), e.data)
This call was effectively like update_data and no longer did anything in bitbake. Drop it as it is obsolete. Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org> --- meta/classes-global/sanity.bbclass | 11 ++--------- 1 file changed, 2 insertions(+), 9 deletions(-)