| Message ID | 20180717195553.9111-1-alex.bennee@linaro.org |
|---|---|
| Headers | show |
| Series | docker fixes (and one tcg test tweak) | expand |
Alex Bennée <alex.bennee@linaro.org> writes: > Hi, > > I've missed the boat for today's rc1 but I'd like to get this merged > before rc2. The new docker.py change is technically new functionality > but I'm counting it as a usability bug fix as it replaces a random > back trace failure with a preemptive failure and message mentioning > binfmt_misc configuration. This would have saved Richard a lot of head > scratching as he tried to setup a powerpc-user setup to test his > setcontext fix (he had a custom binfmt_misc pointing to his src tree). > > Finally we also drop the runcom test. It was cute that it got > resurrected but it is ultimately a pointless test for something I'm > sure no one actually uses. > > There will be a follow-up RFC series after this that cleans-up some of > the rough edges when your host is not an x86_64 box but that series > won't be targeting the 3.0 release. > > : The following patches need review > : patch docker/disable debian powerpc user cross.patch > : patch docker/drop QEMU_TARGET check fallback in EXECUTABLE.patch > : patch docker/Update debootstrap script after Debian migrat.patch > : patch docker/ignore distro versioning of debootstrap.patch > : patch docker/perform basic binfmt_misc validation in docke.patch > : patch tests/tcg remove runcom test.patch Ping? -- Alex Bennée
On Mon, Jul 23, 2018 at 6:03 PM Alex Bennée <alex.bennee@linaro.org> wrote: > > > Alex Bennée <alex.bennee@linaro.org> writes: > > > Hi, > > > > I've missed the boat for today's rc1 but I'd like to get this merged > > before rc2. The new docker.py change is technically new functionality > > but I'm counting it as a usability bug fix as it replaces a random > > back trace failure with a preemptive failure and message mentioning > > binfmt_misc configuration. This would have saved Richard a lot of head > > scratching as he tried to setup a powerpc-user setup to test his > > setcontext fix (he had a custom binfmt_misc pointing to his src tree). > > > > Finally we also drop the runcom test. It was cute that it got > > resurrected but it is ultimately a pointless test for something I'm > > sure no one actually uses. > > > > There will be a follow-up RFC series after this that cleans-up some of > > the rough edges when your host is not an x86_64 box but that series > > won't be targeting the 3.0 release. > > > > : The following patches need review > > : patch docker/disable debian powerpc user cross.patch > > : patch docker/drop QEMU_TARGET check fallback in EXECUTABLE.patch > > : patch docker/Update debootstrap script after Debian migrat.patch > > : patch docker/ignore distro versioning of debootstrap.patch > > : patch docker/perform basic binfmt_misc validation in docke.patch > > : patch tests/tcg remove runcom test.patch > > Ping? I had questions about patch 13 and 17. Otherwise looks good. Maybe you can drop them for now (including patch 9) and send the PULL for the rest. I don't know if we can catch up -rc2 but I guess the testing fixes are fairly safe to sneak in. :) (My development machine is affected by an office power outage, and I was also busy with upgrading patchew.org. Sorry for the late reply !) Fam
Fam Zheng <famz@redhat.com> writes: > On Mon, Jul 23, 2018 at 6:03 PM Alex Bennée <alex.bennee@linaro.org> wrote: >> >> >> Alex Bennée <alex.bennee@linaro.org> writes: >> >> > Hi, >> > >> > I've missed the boat for today's rc1 but I'd like to get this merged >> > before rc2. The new docker.py change is technically new functionality >> > but I'm counting it as a usability bug fix as it replaces a random >> > back trace failure with a preemptive failure and message mentioning >> > binfmt_misc configuration. This would have saved Richard a lot of head >> > scratching as he tried to setup a powerpc-user setup to test his >> > setcontext fix (he had a custom binfmt_misc pointing to his src tree). >> > >> > Finally we also drop the runcom test. It was cute that it got >> > resurrected but it is ultimately a pointless test for something I'm >> > sure no one actually uses. >> > >> > There will be a follow-up RFC series after this that cleans-up some of >> > the rough edges when your host is not an x86_64 box but that series >> > won't be targeting the 3.0 release. >> > >> > : The following patches need review >> > : patch docker/disable debian powerpc user cross.patch >> > : patch docker/drop QEMU_TARGET check fallback in EXECUTABLE.patch >> > : patch docker/Update debootstrap script after Debian migrat.patch >> > : patch docker/ignore distro versioning of debootstrap.patch >> > : patch docker/perform basic binfmt_misc validation in docke.patch >> > : patch tests/tcg remove runcom test.patch >> >> Ping? > > I had questions about patch 13 and 17. Otherwise looks good. > Maybe you can drop them for now (including patch 9) and send the PULL for > the rest. I can drop/replace 13 - but 17 was there for a better reporting of a failure case Peter found. We can certainly expand it and be smarter in future iterations. > I don't know if we can catch up -rc2 but I guess the testing fixes are > fairly safe to sneak in. :) > > (My development machine is affected by > an > office power outage, and I was > also > busy with upgrading patchew.org. Sorry for the > > late reply > !) > > Fam -- Alex Bennée
On Tue, Jul 24, 2018 at 4:56 PM Alex Bennée <alex.bennee@linaro.org> wrote: > > > Fam Zheng <famz@redhat.com> writes: > > > On Mon, Jul 23, 2018 at 6:03 PM Alex Bennée <alex.bennee@linaro.org> wrote: > >> > >> > >> Alex Bennée <alex.bennee@linaro.org> writes: > >> > >> > Hi, > >> > > >> > I've missed the boat for today's rc1 but I'd like to get this merged > >> > before rc2. The new docker.py change is technically new functionality > >> > but I'm counting it as a usability bug fix as it replaces a random > >> > back trace failure with a preemptive failure and message mentioning > >> > binfmt_misc configuration. This would have saved Richard a lot of head > >> > scratching as he tried to setup a powerpc-user setup to test his > >> > setcontext fix (he had a custom binfmt_misc pointing to his src tree). > >> > > >> > Finally we also drop the runcom test. It was cute that it got > >> > resurrected but it is ultimately a pointless test for something I'm > >> > sure no one actually uses. > >> > > >> > There will be a follow-up RFC series after this that cleans-up some of > >> > the rough edges when your host is not an x86_64 box but that series > >> > won't be targeting the 3.0 release. > >> > > >> > : The following patches need review > >> > : patch docker/disable debian powerpc user cross.patch > >> > : patch docker/drop QEMU_TARGET check fallback in EXECUTABLE.patch > >> > : patch docker/Update debootstrap script after Debian migrat.patch > >> > : patch docker/ignore distro versioning of debootstrap.patch > >> > : patch docker/perform basic binfmt_misc validation in docke.patch > >> > : patch tests/tcg remove runcom test.patch > >> > >> Ping? > > > > I had questions about patch 13 and 17. Otherwise looks good. > > Maybe you can drop them for now (including patch 9) and send the PULL for > > the rest. > > I can drop/replace 13 - but 17 was there for a better reporting of a > failure case Peter found. We can certainly expand it and be smarter in > future iterations. Sounds good to me. Fam > > > > I don't know if we can catch up -rc2 but I guess the testing fixes are > > fairly safe to sneak in. :) > > > > (My development machine is affected by > > an > > office power outage, and I was > > also > > busy with upgrading patchew.org. Sorry for the > > > > late reply > > !) > > > > Fam > > > -- > Alex Bennée
Hi, I've missed the boat for today's rc1 but I'd like to get this merged before rc2. The new docker.py change is technically new functionality but I'm counting it as a usability bug fix as it replaces a random back trace failure with a preemptive failure and message mentioning binfmt_misc configuration. This would have saved Richard a lot of head scratching as he tried to setup a powerpc-user setup to test his setcontext fix (he had a custom binfmt_misc pointing to his src tree). Finally we also drop the runcom test. It was cute that it got resurrected but it is ultimately a pointless test for something I'm sure no one actually uses. There will be a follow-up RFC series after this that cleans-up some of the rough edges when your host is not an x86_64 box but that series won't be targeting the 3.0 release. : The following patches need review : patch docker/disable debian powerpc user cross.patch : patch docker/drop QEMU_TARGET check fallback in EXECUTABLE.patch : patch docker/Update debootstrap script after Debian migrat.patch : patch docker/ignore distro versioning of debootstrap.patch : patch docker/perform basic binfmt_misc validation in docke.patch : patch tests/tcg remove runcom test.patch Alex Bennée (17): tests/.gitignore: don't ignore docker tests docker: base debian-tricore on qemu:debian9 docker: par down QEMU_CONFIGURE_OPTS in debian-tricore-cross docker: fail more gracefully on docker.py check docker: split configure_qemu from build_qemu docker: move make check into check_qemu helper docker: gracefully skip check_qemu docker: Makefile.include don't include partial images docker: disable debian-powerpc-user-cross docker: add test-unit runner docker: add expansion for docker-test-FOO to Makefile.include docker: drop QEMU_TARGET check, fallback in EXECUTABLE not set docker: add --hint to docker.py check docker: add commentary to debian-bootstrap.docker docker: ignore distro versioning of debootstrap docker: perform basic binfmt_misc validation in docker.py tests/tcg: remove runcom test Philippe Mathieu-Daudé (1): docker: Update debootstrap script after Debian migration from Alioth to Salsa tests/.gitignore | 1 + tests/docker/Makefile.include | 26 ++- tests/docker/common.rc | 24 ++- tests/docker/docker.py | 40 +++- .../dockerfiles/debian-bootstrap.docker | 1 + tests/docker/dockerfiles/debian-bootstrap.pre | 13 +- .../dockerfiles/debian-tricore-cross.docker | 6 +- tests/docker/test-clang | 2 +- tests/docker/test-debug | 2 +- tests/docker/test-full | 2 +- tests/docker/test-quick | 2 +- tests/docker/test-unit | 21 ++ tests/tcg/i386/Makefile.target | 5 - tests/tcg/i386/README | 3 - tests/tcg/i386/pi_10.com | Bin 54 -> 0 bytes tests/tcg/i386/runcom.c | 192 ------------------ 16 files changed, 116 insertions(+), 224 deletions(-) create mode 100755 tests/docker/test-unit delete mode 100644 tests/tcg/i386/pi_10.com delete mode 100644 tests/tcg/i386/runcom.c