Skip to contentSkip to conversation

bmap-tools: update from 3.4 to 3.5

PATCH Superseded Alex Kiernan 3 participants 10 messages Feb 14, 2019 – Feb 16, 2019

v1 of 3 Superseded Latest is v3 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 bmap-tools: update from 3.4 to 3.5

bmap-tools 3.5 was released August 2018 but has no release tarball, so switch to using the git fetcher. Also pull in the fix for StopIteration being transformed into RuntimeError for Python 3.7.

---

 .../bmap-tools/{bmap-tools_3.4.bb => bmap-tools_3.5.bb}          | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)
 rename meta/recipes-support/bmap-tools/{bmap-tools_3.4.bb => bmap-tools_3.5.bb} (74%)
meta/recipes-support/bmap-tools/bmap-tools_3.5.bb renamed from meta/recipes-support/bmap-tools/bmap-tools_3.4.bb +5 -4
similarity index 74%
rename from meta/recipes-support/bmap-tools/bmap-tools_3.4.bb
rename to meta/recipes-support/bmap-tools/bmap-tools_3.5.bb
index 2aae9a3fe0d8..25d66df623a0 100644
@@ -9,12 +9,13 @@ SECTION = "console/utils"
 LICENSE = "GPLv2"
 LIC_FILES_CHKSUM = "file://COPYING;md5=b234ee4d69f5fce4486a80fdaf4a4263"
 
-SRC_URI = "https://github.com/intel/${BPN}/releases/download/v${PV}/${BP}.tgz"
+SRC_URI = "git://github.com/intel/${BPN}"
 
-SRC_URI[md5sum] = "883420183611aa137cbaf68a91ef0d7e"
-SRC_URI[sha256sum] = "0064891270c180c190f41925b0f6fb9cecac3056f1168d2592dbe90e5f226c0a"
+SRCREV = "db7087b883bf52cbff063ad17a41cc1cbb85104d"
+S = "${WORKDIR}/git"
+PV .= "+git${SRCPV}"
 
-UPSTREAM_CHECK_URI = "https://github.com/intel/${BPN}/releases"
+UPSTREAM_CHECK_GITTAGREGEX = "v(?P<pver>.*)"
 
 RDEPENDS_${PN} = "python3-core python3-compression python3-mmap python3-setuptools python3-fcntl"
 
Signature · 1 line
2.7.4
#1 lore raw
Quoted 1 line · Alex Kiernan

+PV .= "+git${SRCPV}"

This is unnecessary and adds clutter. Just the plain version number is enough.

Alex

#2 lore raw
Quoted 7 lines · Alexander Kanavin
Quoted 1 line · Alex Kiernan

+PV .= "+git${SRCPV}"

This is unnecessary and adds clutter. Just the plain version number is enough.

If I read the original commit message correctly, it is necessary as the version is 3.5 plus a bugfix?

Cheers,

Richard

#3 lore raw
Quoted 5 lines · Richard Purdie
Quoted 3 lines · Alexander Kanavin

On Thu, 14 Feb 2019 at 22:10, Alex Kiernan <[email protected]> wrote:

Quoted 7 lines
Quoted 4 lines
Quoted 1 line

+PV .= "+git${SRCPV}"

This is unnecessary and adds clutter. Just the plain version number is enough.

If I read the original commit message correctly, it is necessary as the version is 3.5 plus a bugfix?

Seems so. Sorry for the noise.

Alex

#4 lore raw
In reply to Richard Purdie
Quoted 5 lines · Richard Purdie
Quoted 3 lines · Alexander Kanavin

On Thu, 14 Feb 2019 at 22:10, Alex Kiernan <[email protected]> wrote:

Quoted 8 lines
Quoted 4 lines
Quoted 1 line

+PV .= "+git${SRCPV}"

This is unnecessary and adds clutter. Just the plain version number is enough.

If I read the original commit message correctly, it is necessary as the version is 3.5 plus a bugfix?

Yes it was 3.5 + a bug fix.

Ignore v2 that I sent!

Signature · 1 line
Alex Kiernan
#5 lore raw
In reply to Alex Kiernan
Quoted 7 lines · Alex Kiernan

bmap-tools 3.5 was released August 2018 but has no release tarball, so
switch to using the git fetcher. Also pull in the fix for StopIteration
being transformed into RuntimeError for Python 3.7.

Unfortunately this triggers a selftest failure:

https://autobuilder.yoctoproject.org/typhoon/#/builders/56/builds/236

(ignore the virtgl piece, that is something else)

oe-selftest -r wic.Wic2.test_bmap_short
and
oe-selftest -r wic.Wic2.test_bmap_long

should reproduce. Might be as simple as a missing RDEPENDS.

Cheers,

Richard

Quoted 40 lines
 .../bmap-tools/{bmap-tools_3.4.bb => bmap-tools_3.5.bb}          | 9
+++++----
 1 file changed, 5 insertions(+), 4 deletions(-)
 rename meta/recipes-support/bmap-tools/{bmap-tools_3.4.bb => bmap-

tools_3.5.bb} (74%)

meta/recipes-support/bmap-tools/bmap-tools_3.5.bb renamed from meta/recipes-support/bmap-tools/bmap-tools_3.4.bb +0 -0
similarity index 74%
rename from meta/recipes-support/bmap-tools/bmap-tools_3.4.bb
rename to meta/recipes-support/bmap-tools/bmap-tools_3.5.bb
index 2aae9a3fe0d8..25d66df623a0 100644
@@ -9,12 +9,13 @@ SECTION = "console/utils"
 LICENSE = "GPLv2"
 LIC_FILES_CHKSUM =

"file://COPYING;md5=b234ee4d69f5fce4486a80fdaf4a4263"

-SRC_URI = "
https://github.com/intel/${BPN}/releases/download/v${PV}/${BP}.tgz"

meta/recipes-support/bmap-tools/bmap-tools_3.5.bb renamed from meta/recipes-support/bmap-tools/bmap-tools_3.4.bb +1 -2
+SRC_URI = "git://github.com/intel/${BPN}"
 
-SRC_URI[md5sum] = "883420183611aa137cbaf68a91ef0d7e"
-SRC_URI[sha256sum] =

"0064891270c180c190f41925b0f6fb9cecac3056f1168d2592dbe90e5f226c0a"

meta/recipes-support/bmap-tools/bmap-tools_3.5.bb renamed from meta/recipes-support/bmap-tools/bmap-tools_3.4.bb +4 -1
+SRCREV = "db7087b883bf52cbff063ad17a41cc1cbb85104d"
+S = "${WORKDIR}/git"
+PV .= "+git${SRCPV}"
 
-UPSTREAM_CHECK_URI = "https://github.com/intel/${BPN}/releases"
+UPSTREAM_CHECK_GITTAGREGEX = "v(?P<pver>.*)"
 
 RDEPENDS_${PN} = "python3-core python3-compression python3-mmap

python3-setuptools python3-fcntl"

Signature · 1 line
2.7.4
#6 lore raw
Quoted 14 lines · Richard Purdie
Quoted 7 lines · Alex Kiernan

bmap-tools 3.5 was released August 2018 but has no release tarball, so
switch to using the git fetcher. Also pull in the fix for StopIteration
being transformed into RuntimeError for Python 3.7.

Unfortunately this triggers a selftest failure:

https://autobuilder.yoctoproject.org/typhoon/#/builders/56/builds/236

Quoted 8 lines

(ignore the virtgl piece, that is something else)

oe-selftest -r wic.Wic2.test_bmap_short
and
oe-selftest -r wic.Wic2.test_bmap_long

should reproduce. Might be as simple as a missing RDEPENDS.

Looks like it's gained a dependency on six. And thanks for the reproducing steps - I've never got to grips with that stuff before.

That said I'm struggling to get the tests to actually pass. I've added in python3-six to RDEPENDS, but it still fails with ModuleNotFoundError: No module named 'six'?

Signature · 1 line
Alex Kiernan
#7 lore raw
Quoted 24 lines · Alex Kiernan
Quoted 13 lines · Richard Purdie
Quoted 7 lines · Alex Kiernan

bmap-tools 3.5 was released August 2018 but has no release tarball, so
switch to using the git fetcher. Also pull in the fix for StopIteration
being transformed into RuntimeError for Python 3.7.

Unfortunately this triggers a selftest failure:

https://autobuilder.yoctoproject.org/typhoon/#/builders/56/builds/236

Quoted 8 lines

(ignore the virtgl piece, that is something else)

oe-selftest -r wic.Wic2.test_bmap_short
and
oe-selftest -r wic.Wic2.test_bmap_long

should reproduce. Might be as simple as a missing RDEPENDS.

Quoted 7 lines

Looks like it's gained a dependency on six. And thanks for the reproducing steps - I've never got to grips with that stuff before.

That said I'm struggling to get the tests to actually pass. I've added
in python3-six to RDEPENDS, but it still fails with ModuleNotFoundError: No module named 'six'?

We appear to have walked into spools of razor wire :/

This is an open bug:

https://bugzilla.yoctoproject.org/show_bug.cgi?id=10113

which I'm struggling to know how to fix.

What does appear to work locally is adding something like:

# We need the native dependency on python3-six at runtime # See bug 10113
EXTRASYSROOTDEPTASK = "" EXTRASYSROOTDEPTASK_class-native = "do_populate_sysroot" do_populate_sysroot[rdeptask] = "${EXTRASYSROOTDEPTASK}"

Whether this is the right thing to do I'm less sure about.

It may be an idea to start introducing these dependencies where we need them to fix bugs, until we can sort the problem properly in native.bbclass.

Cheers,

Richard

#8 lore raw
Quoted 50 lines · Richard Purdie
Quoted 24 lines · Alex Kiernan
Quoted 13 lines · Richard Purdie
Quoted 7 lines · Alex Kiernan

bmap-tools 3.5 was released August 2018 but has no release tarball, so
switch to using the git fetcher. Also pull in the fix for StopIteration
being transformed into RuntimeError for Python 3.7.

Unfortunately this triggers a selftest failure:

https://autobuilder.yoctoproject.org/typhoon/#/builders/56/builds/236

Quoted 8 lines

(ignore the virtgl piece, that is something else)

oe-selftest -r wic.Wic2.test_bmap_short
and
oe-selftest -r wic.Wic2.test_bmap_long

should reproduce. Might be as simple as a missing RDEPENDS.

Quoted 7 lines

Looks like it's gained a dependency on six. And thanks for the reproducing steps - I've never got to grips with that stuff before.

That said I'm struggling to get the tests to actually pass. I've added
in python3-six to RDEPENDS, but it still fails with ModuleNotFoundError: No module named 'six'?

We appear to have walked into spools of razor wire :/

This is an open bug:

https://bugzilla.yoctoproject.org/show_bug.cgi?id=10113

which I'm struggling to know how to fix.

What does appear to work locally is adding something like:

# We need the native dependency on python3-six at runtime # See bug 10113
EXTRASYSROOTDEPTASK = "" EXTRASYSROOTDEPTASK_class-native = "do_populate_sysroot" do_populate_sysroot[rdeptask] = "${EXTRASYSROOTDEPTASK}"

Quoted 6 lines

Whether this is the right thing to do I'm less sure about.

It may be an idea to start introducing these dependencies where we need them to fix bugs, until we can sort the problem properly in native.bbclass.

Thanks, that seems to work... I'd have been here forever trying to find that.

I'll send a v3 with that change in it.

Signature · 1 line
Alex Kiernan
#9 lore raw
Quoted 26 lines · Alex Kiernan
Quoted 16 lines · Richard Purdie

We appear to have walked into spools of razor wire :/

This is an open bug:

https://bugzilla.yoctoproject.org/show_bug.cgi?id=10113

which I'm struggling to know how to fix.

What does appear to work locally is adding something like:

# We need the native dependency on python3-six at runtime # See bug 10113
EXTRASYSROOTDEPTASK = "" EXTRASYSROOTDEPTASK_class-native = "do_populate_sysroot" do_populate_sysroot[rdeptask] = "${EXTRASYSROOTDEPTASK}"

Quoted 7 lines

Whether this is the right thing to do I'm less sure about.

It may be an idea to start introducing these dependencies where we need
them to fix bugs, until we can sort the problem properly in native.bbclass.

Quoted 4 lines

Thanks, that seems to work... I'd have been here forever trying to find that.

I'll send a v3 with that change in it.

Thanks! I poked a bit more at that underlying RDEPENDS problem as it bothers me. Once I fixed the dependency loop detection code in bitbake, the real errors became clear and once two of them were fixed, things started working more as they should.

I've put together a patchset and thrown it at the autobuilder, see what it makes of it.

I'll either merge your patch with or without the hack depending on how that goes...

Cheers,

Richard