| Message ID | 20240502105721.775868-1-peter.hoyes@arm.com |
|---|---|
| State | New |
| Headers | show |
| Series | u-boot: Enable devupstream | expand |
On Thu, 2024-05-02 at 11:57 +0100, Peter Hoyes via lists.openembedded.org wrote: > To enable downstream testing of the bleeding edge of U-Boot, > add devupstream to the recipe, with SRCREV set to automatically > track the latest revision. > > Signed-off-by: Peter Hoyes <peter.hoyes@arm.com> > --- > meta/recipes-bsp/u-boot/u-boot_2024.04.bb | 3 +++ > 1 file changed, 3 insertions(+) > > diff --git a/meta/recipes-bsp/u-boot/u-boot_2024.04.bb > b/meta/recipes-bsp/u-boot/u-boot_2024.04.bb > index b15bcaa818..dc2e2da8ae 100644 > --- a/meta/recipes-bsp/u-boot/u-boot_2024.04.bb > +++ b/meta/recipes-bsp/u-boot/u-boot_2024.04.bb > @@ -3,3 +3,6 @@ require u-boot.inc > > DEPENDS += "bc-native dtc-native python3-pyelftools-native" > > +BBCLASSEXTEND += "devupstream:target" > +SRCREV:class-devupstream = "${AUTOREV}" > +PV:class-devupstream = "dev+git" Doesn't that mean parsing would hit the network by default every time? Cheers, Richard
On 2 May 2024, at 12:02, Richard Purdie via lists.openembedded.org <richard.purdie=linuxfoundation.org@lists.openembedded.org> wrote: >> +BBCLASSEXTEND += "devupstream:target" >> +SRCREV:class-devupstream = "${AUTOREV}" >> +PV:class-devupstream = "dev+git" > > Doesn't that mean parsing would hit the network by default every time? I _believe_ it only hits when the devupstream variation is actually picked. Is this acceptable? Ross
On Thu, 2024-05-02 at 12:55 +0000, Ross Burton wrote: > On 2 May 2024, at 12:02, Richard Purdie via lists.openembedded.org > <richard.purdie=linuxfoundation.org@lists.openembedded.org> wrote: > > > +BBCLASSEXTEND += "devupstream:target" > > > +SRCREV:class-devupstream = "${AUTOREV}" > > > +PV:class-devupstream = "dev+git" > > > > Doesn't that mean parsing would hit the network by default every > > time? > > I _believe_ it only hits when the devupstream variation is actually > picked. Is this acceptable? I believe you're wrong ;-) I applied this patch and the first thing I'd note is that: $ bitbake -p Parsing of 991 .bb files complete (990 cached, 1 parsed). 1969 targets, 50 skipped, 0 masked, 0 errors. $ bitbake -p Parsing of 991 .bb files complete (990 cached, 1 parsed). 1969 targets, 50 skipped, 0 masked, 0 errors. i.e. it is *always* now parsing one recipe, you can guess which. I then tried "bitbake -pDDD" which in the output includes: DEBUG: /media/build/poky/meta/recipes-bsp/u-boot/u-boot_2024.04.bb: Fetcher accessed the network with the command git -c gc.autoDetach=false -c core.pager=cat -c safe.bareRepository=all ls-remote https://source.denx.de/u-boot/u-boot.git so I believe this is accessing the network on every parse. Cheers, Richard
diff --git a/meta/recipes-bsp/u-boot/u-boot_2024.04.bb b/meta/recipes-bsp/u-boot/u-boot_2024.04.bb index b15bcaa818..dc2e2da8ae 100644 --- a/meta/recipes-bsp/u-boot/u-boot_2024.04.bb +++ b/meta/recipes-bsp/u-boot/u-boot_2024.04.bb @@ -3,3 +3,6 @@ require u-boot.inc DEPENDS += "bc-native dtc-native python3-pyelftools-native" +BBCLASSEXTEND += "devupstream:target" +SRCREV:class-devupstream = "${AUTOREV}" +PV:class-devupstream = "dev+git"
To enable downstream testing of the bleeding edge of U-Boot, add devupstream to the recipe, with SRCREV set to automatically track the latest revision. Signed-off-by: Peter Hoyes <peter.hoyes@arm.com> --- meta/recipes-bsp/u-boot/u-boot_2024.04.bb | 3 +++ 1 file changed, 3 insertions(+)