| Message ID | 20220624145039.49929-1-akihiko.odaki@gmail.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | cutils: Introduce bundle mechanism | expand |
On 6/24/22 16:50, Akihiko Odaki wrote: > Developers often run QEMU without installing. The bundle mechanism > allows to look up files which should be present in installation even in > such a situation. > > It is a general mechanism and can find any files located relative > to the installation tree. The build tree must have a new directory, > qemu-bundle, to represent what files the installation tree would > have for reference by the executables. > > Note that this abandons compatibility with Windows older than 8 to use > PathCchSkipRoot(). The extended support for the prior version, 7 ended > more than 2 years ago, and it is unlikely that anyone would like to run > the latest QEMU on such an old system. Thanks, this looks good. Even though technically it adds more lines than it removes (thanks Windows...), it replaces lots of special cases with a single general mechanism, which is nice and future proof. The only change I'd make (which I can do when queueing the patch) is to change the directory name to something "strange" like "=install", so that it's unlikely to have a directory named like that in the $bindir. Any opinions? Paolo > v8: > * Pass absolute paths to get_relocated_path() (Paolo Bonzini) > * Use meson introspection (Paolo Bonzini) > * Drop "qga: Relocate a path emitted in the help text" as it is no longer > relevant for the bundle mechanism. > > v7: Properly fix --firmwarepath (Daniel P. Berrangé) > > v6: Reuse get_relocated_path() in find_bundle() (Paolo Bonzini) > > v5: > * Prefer qemu-bundle if it exists. (Daniel P. Berrangé) > * Check install_blobs option before installing BIOSes (Paolo Bonzini) > * Add common code to set up qemu-bundle to the top level meson.build > (Paolo Bonzini) > > v4: > * Add Daniel P. Berrangé to CC. Hopefully this helps merging his patch: > https://mail.gnu.org/archive/html/qemu-devel/2022-06/msg02276.html > * Rebased to the latest QEMU. > > v3: > * Note that the bundle mechanism is for any files located relative to the > installation tree including but not limited to datadir. (Peter Maydell) > * Fix "bridge" typo (Philippe Mathieu-Daudé) > > v2: Rebased to the latest QEMU. > > Akihiko Odaki (3): > cutils: Introduce bundle mechanism > datadir: Use bundle mechanism > module: Use bundle mechanism > > Paolo Bonzini (1): > tests/vm: do not specify -bios option > > .travis.yml | 2 +- > docs/about/build-platforms.rst | 2 +- > include/qemu/cutils.h | 18 +++++++-- > meson.build | 4 ++ > pc-bios/keymaps/meson.build | 21 +++------- > pc-bios/meson.build | 13 ++----- > scripts/oss-fuzz/build.sh | 2 +- > scripts/symlink-install-tree.py | 37 ++++++++++++++++++ > softmmu/datadir.c | 22 +---------- > tests/qtest/fuzz/fuzz.c | 15 -------- > tests/vm/fedora | 1 - > tests/vm/freebsd | 1 - > tests/vm/netbsd | 1 - > tests/vm/openbsd | 1 - > util/cutils.c | 68 +++++++++++++++++++++++---------- > util/meson.build | 1 + > util/module.c | 1 - > 17 files changed, 117 insertions(+), 93 deletions(-) > create mode 100755 scripts/symlink-install-tree.py >
On 2022/06/25 0:26, Paolo Bonzini wrote: > The only change I'd make (which I can do when queueing the patch) is to > change the directory name to something "strange" like "=install", so > that it's unlikely to have a directory named like that in the $bindir. > > Any opinions? Actually I prepended qemu- to the directory because all executables of QEMU has the prefix. We can safely say QEMU installation will never collide if no file prefixed with qemu- exists in the bindir. If anyone has an idea what should follow "qemu-", please tell us. Regards, Akihiko Odaki