Switch to glib URI parsing code
Subsystems (2)
PATCH 2/9 tests/lcitool/libvirt-ci: Update to the latest master branch
In reply to Thomas HuthWe need the latest fixes for the lcitool to be able to properly update our CentOS docker file to CentOS Stream 9.
--- tests/lcitool/libvirt-ci | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
tests/lcitool/libvirt-ci +1 -1
@@ -1 +1 @@ -Subproject commit 77c800186f34b21be7660750577cc5582a914deb +Subproject commit e25d813a62fb8fa0ee7aa0748cf720b59a187e71
Signature · 1 line
2.44.0
PATCH 5/9 block/gluster: Use URI parsing code from glib
In reply to Thomas HuthSince version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter.
--- block/gluster.c | 63 +++++++++++++++++++++++++++++-------------------- 1 file changed, 37 insertions(+), 26 deletions(-)
block/gluster.c +37 -26
@@ -17,7 +17,6 @@#include "qapi/error.h" #include "qapi/qmp/qdict.h" #include "qapi/qmp/qerror.h" -#include "qemu/uri.h" #include "qemu/error-report.h" #include "qemu/module.h" #include "qemu/option.h"
@@ -289,9 +288,9 @@ static void glfs_clear_preopened(glfs_t *fs) } } -static int parse_volume_options (BlockdevOptionsGluster *gconf,char *path) +static int parse_volume_options (BlockdevOptionsGluster *gconf,const char *path) { -char *p, *q; +const char *p, *q;if (!path) {return -EINVAL ;
@@ -350,12 +349,15 @@ static int qemu_gluster_parse_uri(BlockdevOptionsGluster *gconf, const char *filename) {SocketAddress *gsconf; -URI *uri; -QueryParams *qp =NULL ; +GUri *uri; +GHashTable *qp =NULL ; +int qp_n;bool is_unix = false;int ret =0 ; +const char *uri_scheme, *uri_query, *uri_server; +int uri_port; - uri =uri_parse (filename); + uri =g_uri_parse (filename,G_URI_FLAGS_NONE ,NULL );if (!uri) {return -EINVAL ; }
@@ -364,14 +366,15 @@ static int qemu_gluster_parse_uri(BlockdevOptionsGluster *gconf,QAPI_LIST_PREPEND (gconf->server, gsconf);/* transport */ -if (!uri->scheme || !strcmp (uri->scheme,"gluster" )) { + uri_scheme =g_uri_get_scheme (uri); +if (!uri_scheme || !strcmp (uri_scheme,"gluster" )) { gsconf->type =SOCKET_ADDRESS_TYPE_INET ; - }else if (!strcmp (uri->scheme,"gluster+tcp" )) { + }else if (!strcmp (uri_scheme,"gluster+tcp" )) { gsconf->type =SOCKET_ADDRESS_TYPE_INET ; - }else if (!strcmp (uri->scheme,"gluster+unix" )) { + }else if (!strcmp (uri_scheme,"gluster+unix" )) { gsconf->type =SOCKET_ADDRESS_TYPE_UNIX ; is_unix = true; - }else if (!strcmp (uri->scheme,"gluster+rdma" )) { + }else if (!strcmp (uri_scheme,"gluster+rdma" )) { gsconf->type =SOCKET_ADDRESS_TYPE_INET ;warn_report ("rdma feature is not supported, falling back to tcp" ); }else {
@@ -379,31 +382,39 @@ static int qemu_gluster_parse_uri(BlockdevOptionsGluster *gconf, goto out; } - ret =parse_volume_options (gconf, uri->path); + ret =parse_volume_options (gconf,g_uri_get_path (uri));if (ret <0 ) { goto out; } - qp =query_params_parse (uri->query); -if (qp->n >1 || (is_unix && !qp->n) || (!is_unix && qp->n)) { - ret = -EINVAL ; - goto out; - } - -if (is_unix) { -if (uri->server || uri->port) { + uri_query =g_uri_get_query (uri); +if (uri_query) { + qp =g_uri_parse_params (uri_query,-1 ,"&" ,G_URI_PARAMS_NONE ,NULL ); +if (!qp) { + ret = -EINVAL ; + goto out; + } + qp_n =g_hash_table_size (qp); +if (qp_n >1 || (is_unix && !qp_n) || (!is_unix && qp_n)) { ret = -EINVAL ; goto out; } -if (strcmp (qp->p[0 ].name,"socket" )) { + } + + uri_server =g_uri_get_host (uri); + uri_port =g_uri_get_port (uri); + +if (is_unix) { +char *uri_socket =g_hash_table_lookup (qp,"socket" ); +if (uri_server || uri_port !=-1 || !uri_socket) { ret = -EINVAL ; goto out; } - gsconf->u.q_unix.path =g_strdup (qp->p[0 ].value); + gsconf->u.q_unix.path =g_strdup (uri_socket); }else { - gsconf->u.inet.host =g_strdup (uri->server ? uri->server :"localhost" ); -if (uri->port) { - gsconf->u.inet.port =g_strdup_printf ("%d" , uri->port); + gsconf->u.inet.host =g_strdup (uri_server ? uri_server :"localhost" ); +if (uri_port !=-1 ) { + gsconf->u.inet.port =g_strdup_printf ("%d" , uri_port); }else { gsconf->u.inet.port =g_strdup_printf ("%d" ,GLUSTER_DEFAULT_PORT ); }
@@ -411,9 +422,9 @@ static int qemu_gluster_parse_uri(BlockdevOptionsGluster *gconf, out:if (qp) { -query_params_free (qp); +g_hash_table_destroy (qp); } -uri_free (uri); +g_uri_unref (uri);return ret; }
Signature · 1 line
2.44.0
cover Switch to glib URI parsing code
In the QEMU 9.1 development cycle, we can drop the support for Ubuntu 20.04 and CentOS 8 since the following major versions of these distributions are available since 2 years already.
This allows us to bump the minimum version of glib to 2.66 which comes with a nice set of URI parsing functions. By switching to these parsing functions, we can finally drop our own URI parsing code in util/uri.c.
NB: We also need to update some of the custom runners in our CI environment first (since they still use Ubuntu 20.04).
Thomas Huth (9):
tests: Remove Ubuntu 20.04 container tests/lcitool/libvirt-ci: Update to the latest master branch
tests: Update our CI to use CentOS Stream 9 instead of 8 Bump minimum glib version to v2.66
block/gluster: Use URI parsing code from glib block/nbd: Use URI parsing code from glib block/nfs: Use URI parsing code from glib block/ssh: Use URI parsing code from glib util/uri: Remove the old URI parsing code
meson.build | 16 +-
include/glib-compat.h | 27 +-
include/qemu/uri.h | 99 --
block/gluster.c | 63 +-
block/nbd.c | 66 +-
block/nfs.c | 102 +-
block/ssh.c | 69 +-
qga/commands-posix-ssh.c | 4 +-
util/uri.c | 1466 -----------------
.gitlab-ci.d/buildtest.yml | 16 +-
.gitlab-ci.d/container-core.yml | 4 +-
.../{centos8.docker => centos9.docker} | 34 +-
tests/docker/dockerfiles/ubuntu2004.docker | 157 --
tests/lcitool/libvirt-ci | 2 +-
tests/lcitool/mappings.yml | 20 -
tests/lcitool/refresh | 3 +-
tests/vm/centos | 4 +-
util/meson.build | 2 +-
18 files changed, 200 insertions(+), 1954 deletions(-)
delete mode 100644 include/qemu/uri.h
delete mode 100644 util/uri.c
rename tests/docker/dockerfiles/{centos8.docker => centos9.docker} (82%)
delete mode 100644 tests/docker/dockerfiles/ubuntu2004.dockerSignature · 1 line
2.44.0
PATCH 3/9 tests: Update our CI to use CentOS Stream 9 instead of 8
RHEL 9 (and thus also the derivatives) are available since two years now, so according to QEMU's support policy, we can drop the active support for the previous major version 8 now.
Thus upgrade our CentOS Stream container to major version 9 now.
---
.gitlab-ci.d/buildtest.yml | 16 ++++-----
.gitlab-ci.d/container-core.yml | 4 +--
.../{centos8.docker => centos9.docker} | 34 +++++++------------
tests/lcitool/mappings.yml | 20 -----------
tests/lcitool/refresh | 2 +-
tests/vm/centos | 4 +--
6 files changed, 26 insertions(+), 54 deletions(-)
rename tests/docker/dockerfiles/{centos8.docker => centos9.docker} (82%).gitlab-ci.d/buildtest.yml +8 -8
@@ -158,9 +158,9 @@ build-system-centos: - .native_build_job_template - .native_build_artifact_template needs: - job: amd64-centos8-container + job: amd64-centos9-container variables: - IMAGE: centos8 + IMAGE: centos9 CONFIGURE_ARGS: --disable-nettle --enable-gcrypt --enable-vfio-user-server --enable-modules --enable-trace-backends=dtrace --enable-docs TARGETS: ppc64-softmmu or1k-softmmu s390x-softmmu
@@ -242,7 +242,7 @@ check-system-centos: - job: build-system-centos artifacts: true variables: - IMAGE: centos8 + IMAGE: centos9 MAKE_CHECK_ARGS: check avocado-system-centos:
@@ -251,7 +251,7 @@ avocado-system-centos: - job: build-system-centos artifacts: true variables: - IMAGE: centos8 + IMAGE: centos9 MAKE_CHECK_ARGS: check-avocado AVOCADO_TAGS: arch:ppc64 arch:or1k arch:s390x arch:x86_64 arch:rx arch:sh4 arch:nios2
@@ -327,9 +327,9 @@ avocado-system-flaky: build-tcg-disabled: extends: .native_build_job_template needs: - job: amd64-centos8-container + job: amd64-centos9-container variables: - IMAGE: centos8 + IMAGE: centos9 script: - mkdir build - cd build
@@ -651,9 +651,9 @@ build-tci: build-without-defaults: extends: .native_build_job_template needs: - job: amd64-centos8-container + job: amd64-centos9-container variables: - IMAGE: centos8 + IMAGE: centos9 CONFIGURE_ARGS: --without-default-devices --without-default-features
.gitlab-ci.d/container-core.yml +2 -2
@@ -1,10 +1,10 @@ include: - local: '/.gitlab-ci.d/container-template.yml' -amd64-centos8-container: +amd64-centos9-container: extends: .container_job_template variables: - NAME: centos8 + NAME: centos9 amd64-fedora-container: extends: .container_job_template
tests/docker/dockerfiles/centos9.docker renamed from tests/docker/dockerfiles/centos8.docker +13 -21
@@ -1,15 +1,14 @@ # THIS FILE WAS AUTO-GENERATED # -# $ lcitool dockerfile --layers all centos-stream-8 qemu +# $ lcitool dockerfile --layers all centos-stream-9 qemu # # https://gitlab.com/libvirt/libvirt-ci -FROM quay.io/centos/centos:stream8 +FROM quay.io/centos/centos:stream9 RUN dnf distro-sync -y && \ dnf install 'dnf-command(config-manager)' -y && \ - dnf config-manager --set-enabled -y powertools && \ - dnf install -y centos-release-advanced-virtualization && \ + dnf config-manager --set-enabled -y crb && \ dnf install -y epel-release && \ dnf install -y epel-next-release && \ dnf install -y \
@@ -42,7 +41,6 @@ RUN dnf distro-sync -y && \ glib2-static \ glibc-langpack-en \ glibc-static \ - glusterfs-api-devel \ gnutls-devel \ gtk3-devel \ hostname \
@@ -82,6 +80,7 @@ RUN dnf distro-sync -y && \ lzo-devel \ make \ mesa-libgbm-devel \ + meson \ mtools \ ncurses-devel \ nettle-devel \
@@ -95,25 +94,25 @@ RUN dnf distro-sync -y && \ pixman-devel \ pkgconfig \ pulseaudio-libs-devel \ - python38 \ - python38-PyYAML \ - python38-numpy \ - python38-pip \ - python38-setuptools \ - python38-wheel \ + python3 \ + python3-PyYAML \ + python3-numpy \ + python3-pillow \ + python3-pip \ + python3-sphinx \ + python3-sphinx_rtd_theme \ + python3-tomli \ rdma-core-devel \ sed \ snappy-devel \ socat \ spice-protocol \ - spice-server-devel \ swtpm \ systemd-devel \ systemtap-sdt-devel \ tar \ usbredir-devel \ util-linux \ - virglrenderer-devel \ vte291-devel \ which \ xfsprogs-devel \
@@ -131,18 +130,11 @@ RUN dnf distro-sync -y && \ ln -s /usr/bin/ccache /usr/libexec/ccache-wrappers/g++ && \ ln -s /usr/bin/ccache /usr/libexec/ccache-wrappers/gcc -RUN /usr/bin/pip3.8 install \ - meson==0.63.2 \ - pillow \ - sphinx \ - sphinx-rtd-theme \ - tomli - ENV CCACHE_WRAPPERSDIR "/usr/libexec/ccache-wrappers" ENV LANG "en_US.UTF-8" ENV MAKE "/usr/bin/make" ENV NINJA "/usr/bin/ninja" -ENV PYTHON "/usr/bin/python3.8" +ENV PYTHON "/usr/bin/python3" # As a final step configure the user (if env is defined) ARG USER ARG UID
tests/lcitool/mappings.yml +0 -20
@@ -1,66 +1,50 @@ mappings: flake8: - CentOSStream8: OpenSUSELeap15: meson: - CentOSStream8: OpenSUSELeap15: python3: - CentOSStream8: python38 OpenSUSELeap15: python311-base python3-PyYAML: - CentOSStream8: python38-PyYAML OpenSUSELeap15: python3-devel: - CentOSStream8: python38-devel OpenSUSELeap15: python311-devel python3-docutils: - CentOSStream8: OpenSUSELeap15: python3-numpy: - CentOSStream8: python38-numpy OpenSUSELeap15: python3-opencv: - CentOSStream8: OpenSUSELeap15: python3-pillow: - CentOSStream8: OpenSUSELeap15: python3-pip: - CentOSStream8: python38-pip OpenSUSELeap15: python311-pip python3-pillow: - CentOSStream8: OpenSUSELeap15: python3-selinux: - CentOSStream8: OpenSUSELeap15: python3-setuptools: - CentOSStream8: python38-setuptools OpenSUSELeap15: python311-setuptools python3-sphinx: - CentOSStream8: OpenSUSELeap15: python3-sphinx-rtd-theme: - CentOSStream8: OpenSUSELeap15: python3-sqlite3: - CentOSStream8: python38 OpenSUSELeap15: python311 python3-tomli:
@@ -69,15 +53,11 @@ mappings: Fedora: Debian12: OpenSUSELeap15: - # Not available for Python 3.8 - CentOSStream8: python3-venv: - CentOSStream8: python38 OpenSUSELeap15: python311-base python3-wheel: - CentOSStream8: python38-wheel OpenSUSELeap15: python311-pip pypi_mappings:
tests/lcitool/refresh +1 -1
@@ -125,7 +125,7 @@ try: # Standard native builds # generate_dockerfile("alpine", "alpine-318") - generate_dockerfile("centos8", "centos-stream-8") + generate_dockerfile("centos9", "centos-stream-9") generate_dockerfile("debian", "debian-12", trailer="".join(debian12_extras)) generate_dockerfile("fedora", "fedora-38")
tests/vm/centos +2 -2
@@ -26,8 +26,8 @@ class CentosVM(basevm.BaseVM): export SRC_ARCHIVE=/dev/vdb; sudo chmod a+r $SRC_ARCHIVE; tar -xf $SRC_ARCHIVE; - make docker-test-block@centos8 {verbose} J={jobs} NETWORK=1; - make docker-test-quick@centos8 {verbose} J={jobs} NETWORK=1; + make docker-test-block@centos9 {verbose} J={jobs} NETWORK=1; + make docker-test-quick@centos9 {verbose} J={jobs} NETWORK=1; """ def build_image(self, img):
Signature · 1 line
2.44.0
PATCH 4/9 Bump minimum glib version to v2.66
In reply to Thomas HuthNow that we dropped support for CentOS 8 and Ubuntu 20.04, we can look into bumping the glib version to a new minimum for further clean-ups. According to repology.org, available versions are:
CentOS Stream 9: 2.66.7 Debian 11: 2.66.8 Fedora 38: 2.74.1 Freebsd: 2.78.4 Homebrew: 2.80.0 Openbsd: 2.78.4 OpenSuse leap 15.5: 2.70.5 pkgsrc_current: 2.78.4 Ubuntu 22.04: 2.72.1
Thus it should be safe to bump the minimum glib version to 2.66 now. Version 2.66 comes with new functions for URI parsing which will allow further clean-ups in the following patches.
--- meson.build | 16 +--------------- include/glib-compat.h | 27 ++------------------------- qga/commands-posix-ssh.c | 4 ++-- 3 files changed, 5 insertions(+), 42 deletions(-)
meson.build +1 -15
@@ -865,7 +865,7 @@ have_xen_pci_passthrough = get_option('xen_pci_passthrough') \ # When bumping glib minimum version, please check also whether to increase # the _WIN32_WINNT setting in osdep.h according to the value from glib -glib_req_ver = '>=2.56.0' +glib_req_ver = '>=2.66.0' glib_pc = dependency('glib-2.0', version: glib_req_ver, required: true, method: 'pkg-config') glib_cflags = []
@@ -906,20 +906,6 @@ if not cc.compiles(''' to the right pkg-config files for your build target.''') endif -# Silence clang warnings triggered by glib < 2.57.2 -if not cc.compiles(''' - #include <glib.h> - typedef struct Foo { - int i; - } Foo; - static void foo_free(Foo *f) - { - g_free(f); - } - G_DEFINE_AUTOPTR_CLEANUP_FUNC(Foo, foo_free) - int main(void) { return 0; }''', dependencies: glib_pc, args: ['-Wunused-function', '-Werror']) - glib_cflags += cc.get_supported_arguments('-Wno-unused-function') -endif glib = declare_dependency(dependencies: [glib_pc, gmodule], compile_args: glib_cflags, version: glib_pc.version())
include/glib-compat.h +2 -25
@@ -19,12 +19,12 @@/* Ask for warnings for anything that was marked deprecated in * the defined version, or before. It is a candidate for rewrite. */ -#define GLIB_VERSION_MIN_REQUIRED GLIB_VERSION_2_56 +#define GLIB_VERSION_MIN_REQUIRED GLIB_VERSION_2_66/* Ask for warnings if code tries to use function that did not * exist in the defined version. These risk breaking builds */ -#define GLIB_VERSION_MAX_ALLOWED GLIB_VERSION_2_56 +#define GLIB_VERSION_MAX_ALLOWED GLIB_VERSION_2_66#pragma GCC diagnostic push#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
@@ -105,29 +105,6 @@ static inline gpointer g_memdup2_qemu(gconstpointer mem, gsize byte_size) }#define g_memdup2 (m, s) g_memdup2_qemu(m, s) -#if defined(G_OS_UNIX ) -/* -* Note: The fallback implementation is not MT-safe, and it returns a copy of -* the libc passwd (must be g_free() after use) but not the content. Because of -* these important differences the caller must be aware of, it's not #define for -* GLib API substitution. -*/ -static inline struct passwd * -g_unix_get_passwd_entry_qemu (const gchar *user_name,GError **error) -{ -#if GLIB_CHECK_VERSION (2 ,64 ,0 ) -return g_unix_get_passwd_entry (user_name, error); -#else -struct passwd *p =getpwnam (user_name); -if (!p) { -g_set_error_literal (error,G_UNIX_ERROR ,0 ,g_strerror (errno)); -return NULL ; - } -return (struct passwd *)g_memdup (p,sizeof (*p)); -#endif -} -#endif /* G_OS_UNIX */ -static inline bool qemu_g_test_slow (void ) {
qga/commands-posix-ssh.c +2 -2
@@ -35,7 +35,7 @@ test_get_passwd_entry(const gchar *user_name, GError **error) return p; } -#define g_unix_get_passwd_entry_qemu (username, err) \ +#define g_unix_get_passwd_entry (username, err) \ test_get_passwd_entry(username, err)#endif
@@ -45,7 +45,7 @@ get_passwd_entry(const char *username, Error **errp)g_autoptr (GError) err =NULL ;struct passwd *p; - p =g_unix_get_passwd_entry_qemu (username, &err); + p =g_unix_get_passwd_entry (username, &err);if (p ==NULL ) {error_setg (errp,"failed to lookup user '%s': %s" , username, err->message);
Signature · 1 line
2.44.0
PATCH 1/9 tests: Remove Ubuntu 20.04 container
In reply to Thomas HuthSince Ubuntu 22.04 is now available since two years, we can stop actively supporting the previous LTS version of Ubuntu now.
--- tests/docker/dockerfiles/ubuntu2004.docker | 157 --------------------- tests/lcitool/refresh | 1 - 2 files changed, 158 deletions(-) delete mode 100644 tests/docker/dockerfiles/ubuntu2004.docker
tests/docker/dockerfiles/ubuntu2004.docker deleted +0 -157
@@ -1,157 +0,0 @@ -# THIS FILE WAS AUTO-GENERATED -# -# $ lcitool dockerfile --layers all ubuntu-2004 qemu -# -# https://gitlab.com/libvirt/libvirt-ci - -FROM docker.io/library/ubuntu:20.04 - -RUN export DEBIAN_FRONTEND=noninteractive && \ - apt-get update && \ - apt-get install -y eatmydata && \ - eatmydata apt-get dist-upgrade -y && \ - eatmydata apt-get install --no-install-recommends -y \ - bash \ - bc \ - bison \ - bsdmainutils \ - bzip2 \ - ca-certificates \ - ccache \ - clang \ - dbus \ - debianutils \ - diffutils \ - exuberant-ctags \ - findutils \ - flex \ - g++ \ - gcc \ - gcovr \ - gettext \ - git \ - hostname \ - libaio-dev \ - libasan6 \ - libasound2-dev \ - libattr1-dev \ - libbrlapi-dev \ - libbz2-dev \ - libc6-dev \ - libcacard-dev \ - libcap-ng-dev \ - libcapstone-dev \ - libcmocka-dev \ - libcurl4-gnutls-dev \ - libdaxctl-dev \ - libdrm-dev \ - libepoxy-dev \ - libfdt-dev \ - libffi-dev \ - libfuse3-dev \ - libgbm-dev \ - libgcrypt20-dev \ - libglib2.0-dev \ - libglusterfs-dev \ - libgnutls28-dev \ - libgtk-3-dev \ - libibumad-dev \ - libibverbs-dev \ - libiscsi-dev \ - libjemalloc-dev \ - libjpeg-turbo8-dev \ - libjson-c-dev \ - liblttng-ust-dev \ - liblzo2-dev \ - libncursesw5-dev \ - libnfs-dev \ - libnuma-dev \ - libpam0g-dev \ - libpcre2-dev \ - libpixman-1-dev \ - libpmem-dev \ - libpng-dev \ - libpulse-dev \ - librbd-dev \ - librdmacm-dev \ - libsasl2-dev \ - libsdl2-dev \ - libsdl2-image-dev \ - libseccomp-dev \ - libselinux1-dev \ - libslirp-dev \ - libsnappy-dev \ - libsndio-dev \ - libspice-protocol-dev \ - libspice-server-dev \ - libssh-dev \ - libsystemd-dev \ - libtasn1-6-dev \ - libubsan1 \ - libudev-dev \ - libusb-1.0-0-dev \ - libusbredirhost-dev \ - libvdeplug-dev \ - libvirglrenderer-dev \ - libvte-2.91-dev \ - libxen-dev \ - libzstd-dev \ - llvm \ - locales \ - make \ - mtools \ - multipath-tools \ - ncat \ - nettle-dev \ - ninja-build \ - openssh-client \ - pkgconf \ - python3 \ - python3-numpy \ - python3-opencv \ - python3-pillow \ - python3-pip \ - python3-setuptools \ - python3-sphinx \ - python3-sphinx-rtd-theme \ - python3-venv \ - python3-wheel \ - python3-yaml \ - rpm2cpio \ - sed \ - socat \ - sparse \ - systemtap-sdt-dev \ - tar \ - tesseract-ocr \ - tesseract-ocr-eng \ - xfslibs-dev \ - xorriso \ - zlib1g-dev \ - zstd && \ - eatmydata apt-get autoremove -y && \ - eatmydata apt-get autoclean -y && \ - sed -Ei 's,^# (en_US\.UTF-8 .*)$,\1,' /etc/locale.gen && \ - dpkg-reconfigure locales && \ - dpkg-query --showformat '${Package}_${Version}_${Architecture}\n' --show > /packages.txt && \ - mkdir -p /usr/libexec/ccache-wrappers && \ - ln -s /usr/bin/ccache /usr/libexec/ccache-wrappers/c++ && \ - ln -s /usr/bin/ccache /usr/libexec/ccache-wrappers/cc && \ - ln -s /usr/bin/ccache /usr/libexec/ccache-wrappers/clang && \ - ln -s /usr/bin/ccache /usr/libexec/ccache-wrappers/g++ && \ - ln -s /usr/bin/ccache /usr/libexec/ccache-wrappers/gcc - -RUN /usr/bin/pip3 install \ - meson==0.63.2 \ - tomli - -ENV CCACHE_WRAPPERSDIR "/usr/libexec/ccache-wrappers" -ENV LANG "en_US.UTF-8" -ENV MAKE "/usr/bin/make" -ENV NINJA "/usr/bin/ninja" -ENV PYTHON "/usr/bin/python3" -# As a final step configure the user (if env is defined) -ARG USER -ARG UID -RUN if [ "${USER}" ]; then \ - id ${USER} 2>/dev/null || useradd -u ${UID} -U ${USER}; fi
tests/lcitool/refresh +0 -1
@@ -130,7 +130,6 @@ try: trailer="".join(debian12_extras)) generate_dockerfile("fedora", "fedora-38") generate_dockerfile("opensuse-leap", "opensuse-leap-15") - generate_dockerfile("ubuntu2004", "ubuntu-2004") generate_dockerfile("ubuntu2204", "ubuntu-2204") #
Signature · 1 line
2.44.0
PATCH 9/9 util/uri: Remove the old URI parsing code
In reply to Thomas HuthNow that we switched all consumers of the URI code to use the URI parsing functions from glib instead, we can remove our internal URI parsing code since it is not used anymore.
--- include/qemu/uri.h | 99 --- util/uri.c | 1466 -------------------------------------------- util/meson.build | 2 +- 3 files changed, 1 insertion(+), 1566 deletions(-) delete mode 100644 include/qemu/uri.h delete mode 100644 util/uri.c
include/qemu/uri.h deleted +0 -99
@@ -1,99 +0,0 @@ -/** -* Summary: library of generic URI related routines -* Description: library of generic URI related routines -* Implements RFC 2396 -* -* Copyright (C) 1998-2003 Daniel Veillard. All Rights Reserved. -* -* Permission is hereby granted, free of charge, to any person obtaining a copy -* of this software and associated documentation files (the "Software"), to deal -* in the Software without restriction, including without limitation the rights -* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell -* copies of the Software, and to permit persons to whom the Software is -* furnished to do so, subject to the following conditions: -* -* The above copyright notice and this permission notice shall be included in -* all copies or substantial portions of the Software. -* -* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR -* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, -* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE -* DANIEL VEILLARD BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER -* IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN -* CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. -* -* Except as contained in this notice, the name of Daniel Veillard shall not -* be used in advertising or otherwise to promote the sale, use or other -* dealings in this Software without prior written authorization from him. -* -* Author: Daniel Veillard -** -* Copyright (C) 2007 Red Hat, Inc. -* -* This library is free software; you can redistribute it and/or -* modify it under the terms of the GNU Lesser General Public -* License as published by the Free Software Foundation; either -* version 2.1 of the License, or (at your option) any later version. -* -* This library is distributed in the hope that it will be useful, -* but WITHOUT ANY WARRANTY; without even the implied warranty of -* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU -* Lesser General Public License for more details. -* -* You should have received a copy of the GNU Lesser General Public -* License along with this library. If not, see <https://www.gnu.org/licenses/>. -* -* Authors: -* Richard W.M. Jones <[email protected]> -* -* Utility functions to help parse and assemble query strings. -*/ - -#ifndef QEMU_URI_H -#define QEMU_URI_H - -/** -* URI: -* -* A parsed URI reference. This is a struct containing the various fields -* as described in RFC 2396 but separated for further processing. -*/ -typedef struct URI { -char *scheme;/* the URI scheme */ -char *opaque;/* opaque part */ -char *authority;/* the authority part */ -char *server;/* the server part */ -char *user;/* the user part */ -int port;/* the port number */ -char *path;/* the path string */ -char *fragment;/* the fragment identifier */ -int cleanup;/* parsing potentially unclean URI */ -char *query;/* the query string (as it appears in the URI) */ -}URI ; - -URI *uri_new (void ); -URI *uri_parse (const char *str); -URI *uri_parse_raw (const char *str,int raw); -int uri_parse_into (URI *uri,const char *str); -char *uri_to_string (URI *uri); -void uri_free (URI *uri); - -/* Single web service query parameter 'name=value'. */ -typedef struct QueryParam { -char *name;/* Name (unescaped). */ -char *value;/* Value (unescaped). */ -int ignore;/* Ignore this field in qparam_get_query */ -}QueryParam ; - -/* Set of parameters. */ -typedef struct QueryParams { -int n;/* number of parameters used */ -int alloc;/* allocated space */ -QueryParam *p;/* array of parameters */ -}QueryParams ; - -QueryParams *query_params_new (int init_alloc); -QueryParams *query_params_parse (const char *query); -void query_params_free (QueryParams *ps); - -#endif /* QEMU_URI_H */
util/uri.c deleted +0 -1466
1,466 lines not on this page · raw message
util/meson.build +1 -1
@@ -94,7 +94,7 @@ if have_block util_ss.add(files('hbitmap.c')) util_ss.add(files('hexdump.c')) util_ss.add(files('iova-tree.c')) - util_ss.add(files('iov.c', 'uri.c')) + util_ss.add(files('iov.c')) util_ss.add(files('nvdimm-utils.c')) util_ss.add(files('block-helpers.c')) util_ss.add(files('qemu-coroutine-sleep.c'))
Signature · 1 line
2.44.0
PATCH 6/9 block/nbd: Use URI parsing code from glib
In reply to Thomas HuthSince version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter. The g_uri_get_host() also takes care of removing the square brackets from IPv6 addresses, so we can drop that part of the QEMU code now, too.
--- block/nbd.c | 66 ++++++++++++++++++++++++++++++----------------------- 1 file changed, 38 insertions(+), 28 deletions(-)
block/nbd.c +38 -28
@@ -31,7 +31,6 @@#include "qemu/osdep.h" #include "trace.h" -#include "qemu/uri.h" #include "qemu/option.h" #include "qemu/cutils.h" #include "qemu/main-loop.h"
@@ -1514,30 +1513,34 @@ static void nbd_client_close(BlockDriverState *bs)static int nbd_parse_uri (const char *filename,QDict *options) { -URI *uri; +GUri *uri;const char *p; -QueryParams *qp =NULL ; +GHashTable *qp =NULL ; +int qp_n;int ret =0 ;bool is_unix; +const char *uri_scheme, *uri_query, *uri_server; +int uri_port; - uri =uri_parse (filename); + uri =g_uri_parse (filename,G_URI_FLAGS_NONE ,NULL );if (!uri) {return -EINVAL ; }/* transport */ -if (!g_strcmp0 (uri->scheme,"nbd" )) { + uri_scheme =g_uri_get_scheme (uri); +if (!g_strcmp0 (uri_scheme,"nbd" )) { is_unix = false; - }else if (!g_strcmp0 (uri->scheme,"nbd+tcp" )) { + }else if (!g_strcmp0 (uri_scheme,"nbd+tcp" )) { is_unix = false; - }else if (!g_strcmp0 (uri->scheme,"nbd+unix" )) { + }else if (!g_strcmp0 (uri_scheme,"nbd+unix" )) { is_unix = true; }else { ret = -EINVAL ; goto out; } - p = uri->path ? uri->path :"" ; + p =g_uri_get_path (uri) ?:"" ;if (p[0 ] =='/' ) { p++; }
@@ -1545,51 +1548,58 @@ static int nbd_parse_uri(const char *filename, QDict *options)qdict_put_str (options,"export" , p); } - qp =query_params_parse (uri->query); -if (qp->n >1 || (is_unix && !qp->n) || (!is_unix && qp->n)) { - ret = -EINVAL ; - goto out; + uri_query =g_uri_get_query (uri); +if (uri_query) { + qp =g_uri_parse_params (uri_query,-1 ,"&" ,G_URI_PARAMS_NONE ,NULL ); +if (!qp) { + ret = -EINVAL ; + goto out; + } + qp_n =g_hash_table_size (qp); +if (qp_n >1 || (is_unix && !qp_n) || (!is_unix && qp_n)) { + ret = -EINVAL ; + goto out; + } + } + + uri_server =g_uri_get_host (uri); +if (uri_server && !uri_server[0 ]) { + uri_server =NULL ; } + uri_port =g_uri_get_port (uri);if (is_unix) {/* nbd+unix:///export?socket=path */ -if (uri->server || uri->port ||strcmp (qp->p[0 ].name,"socket" )) { +const char *uri_socket =g_hash_table_lookup (qp,"socket" ); +if (uri_server || uri_port !=-1 || !uri_socket) { ret = -EINVAL ; goto out; }qdict_put_str (options,"server.type" ,"unix" ); -qdict_put_str (options,"server.path" , qp->p[0 ].value); +qdict_put_str (options,"server.path" , uri_socket); }else { -QString *host;char *port_str;/* nbd[+tcp]://host[:port]/export */ -if (!uri->server) { +if (!uri_server) { ret = -EINVAL ; goto out; } -/* strip braces from literal IPv6 address */ -if (uri->server[0 ] =='[' ) { - host =qstring_from_substr (uri->server,1 , -strlen (uri->server) -1 ); - }else { - host =qstring_from_str (uri->server); - } -qdict_put_str (options,"server.type" ,"inet" ); -qdict_put (options,"server.host" , host); +qdict_put_str (options,"server.host" , uri_server); - port_str =g_strdup_printf ("%d" , uri->port ?:NBD_DEFAULT_PORT ); + port_str =g_strdup_printf ("%d" , uri_port !=-1 ? uri_port + :NBD_DEFAULT_PORT );qdict_put_str (options,"server.port" , port_str);g_free (port_str); } out:if (qp) { -query_params_free (qp); +g_hash_table_destroy (qp); } -uri_free (uri); +g_uri_unref (uri);return ret; }
Signature · 1 line
2.44.0
PATCH 7/9 block/nfs: Use URI parsing code from glib
In reply to Thomas HuthSince version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter.
--- block/nfs.c | 102 ++++++++++++++++++++++++++-------------------------- 1 file changed, 52 insertions(+), 50 deletions(-)
block/nfs.c +52 -50
@@ -38,7 +38,6 @@#include "qemu/main-loop.h" #include "qemu/module.h" #include "qemu/option.h" -#include "qemu/uri.h" #include "qemu/cutils.h" #include "sysemu/replay.h" #include "qapi/qapi-visit-block-core.h"
@@ -79,76 +78,79 @@ typedef struct NFSRPC {static int nfs_parse_uri (const char *filename,QDict *options,Error **errp) { -URI *uri =NULL ; -QueryParams *qp =NULL ; -int ret = -EINVAL , i; - - uri =uri_parse (filename); +GUri *uri; +GUriParamsIter qp; +const char *uri_server, *uri_path, *uri_query; +char *qp_name, *qp_value; +GError *gerror =NULL ; +int ret = -EINVAL ; + + uri =g_uri_parse (filename,G_URI_FLAGS_NONE ,NULL );if (!uri) {error_setg (errp,"Invalid URI specified" ); goto out; } -if (g_strcmp0 (uri->scheme,"nfs" ) !=0 ) { +if (!g_str_equal (g_uri_get_scheme (uri),"nfs" )) {error_setg (errp,"URI scheme must be 'nfs'" ); goto out; } -if (!uri->server) { + uri_server =g_uri_get_host (uri); +if (!uri_server || !uri_server[0 ]) {error_setg (errp,"missing hostname in URI" ); goto out; } -if (!uri->path) { + uri_path =g_uri_get_path (uri); +if (!uri_path || !uri_path[0 ]) {error_setg (errp,"missing file path in URI" ); goto out; } - qp =query_params_parse (uri->query); -if (!qp) { -error_setg (errp,"could not parse query parameters" ); - goto out; - } - -qdict_put_str (options,"server.host" , uri->server); +qdict_put_str (options,"server.host" , uri_server);qdict_put_str (options,"server.type" ,"inet" ); -qdict_put_str (options,"path" , uri->path); - -for (i =0 ; i < qp->n; i++) { -uint64_t val; -if (!qp->p[i].value) { -error_setg (errp,"Value for NFS parameter expected: %s" , - qp->p[i].name); - goto out; - } -if (parse_uint_full (qp->p[i].value,0 , &val)) { -error_setg (errp,"Illegal value for NFS parameter: %s" , - qp->p[i].name); - goto out; - } -if (!strcmp (qp->p[i].name,"uid" )) { -qdict_put_str (options,"user" , qp->p[i].value); - }else if (!strcmp (qp->p[i].name,"gid" )) { -qdict_put_str (options,"group" , qp->p[i].value); - }else if (!strcmp (qp->p[i].name,"tcp-syncnt" )) { -qdict_put_str (options,"tcp-syn-count" , qp->p[i].value); - }else if (!strcmp (qp->p[i].name,"readahead" )) { -qdict_put_str (options,"readahead-size" , qp->p[i].value); - }else if (!strcmp (qp->p[i].name,"pagecache" )) { -qdict_put_str (options,"page-cache-size" , qp->p[i].value); - }else if (!strcmp (qp->p[i].name,"debug" )) { -qdict_put_str (options,"debug" , qp->p[i].value); - }else { -error_setg (errp,"Unknown NFS parameter name: %s" , - qp->p[i].name); - goto out; +qdict_put_str (options,"path" , uri_path); + + uri_query =g_uri_get_query (uri); +if (uri_query) { +g_uri_params_iter_init (&qp, uri_query,-1 ,"&" ,G_URI_PARAMS_NONE ); +while (g_uri_params_iter_next (&qp, &qp_name, &qp_value, &gerror)) { +uint64_t val; +if (!qp_name || gerror) { +error_setg (errp,"Failed to parse NFS parameter" ); + goto out; + } +if (!qp_value) { +error_setg (errp,"Value for NFS parameter expected: %s" , + qp_name); + goto out; + } +if (parse_uint_full (qp_value,0 , &val)) { +error_setg (errp,"Illegal value for NFS parameter: %s" , + qp_name); + goto out; + } +if (g_str_equal (qp_name,"uid" )) { +qdict_put_str (options,"user" , qp_value); + }else if (g_str_equal (qp_name,"gid" )) { +qdict_put_str (options,"group" , qp_value); + }else if (g_str_equal (qp_name,"tcp-syncnt" )) { +qdict_put_str (options,"tcp-syn-count" , qp_value); + }else if (g_str_equal (qp_name,"readahead" )) { +qdict_put_str (options,"readahead-size" , qp_value); + }else if (g_str_equal (qp_name,"pagecache" )) { +qdict_put_str (options,"page-cache-size" , qp_value); + }else if (g_str_equal (qp_name,"debug" )) { +qdict_put_str (options,"debug" , qp_value); + }else { +error_setg (errp,"Unknown NFS parameter name: %s" , qp_name); + goto out; + } } } ret =0 ; out: -if (qp) { -query_params_free (qp); - } -uri_free (uri); +g_uri_unref (uri);return ret; }
Signature · 1 line
2.44.0
PATCH 8/9 block/ssh: Use URI parsing code from glib
In reply to Thomas HuthSince version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter.
--- block/ssh.c | 69 +++++++++++++++++++++++++++++++---------------------- 1 file changed, 40 insertions(+), 29 deletions(-)
block/ssh.c +40 -29
@@ -37,7 +37,6 @@#include "qemu/ctype.h" #include "qemu/cutils.h" #include "qemu/sockets.h" -#include "qemu/uri.h" #include "qapi/qapi-visit-sockets.h" #include "qapi/qapi-visit-block-core.h" #include "qapi/qmp/qdict.h"
@@ -181,64 +180,76 @@ static void sftp_error_trace(BDRVSSHState *s, const char *op)static int parse_uri (const char *filename,QDict *options,Error **errp) { -URI *uri =NULL ; -QueryParams *qp; +GUri *uri; +const char *uri_host, *uri_path, *uri_user, *uri_query;char *port_str; -int i; +int port; +g_autoptr (GError) gerror =NULL ; +char *qp_name, *qp_value; +GUriParamsIter qp; - uri =uri_parse (filename); + uri =g_uri_parse (filename,G_URI_FLAGS_NONE ,NULL );if (!uri) {return -EINVAL ; } -if (g_strcmp0 (uri->scheme,"ssh" ) !=0 ) { +if (g_strcmp0 (g_uri_get_scheme (uri),"ssh" ) !=0 ) {error_setg (errp,"URI scheme must be 'ssh'" ); goto err; } -if (!uri->server ||strcmp (uri->server,"" ) ==0 ) { + uri_host =g_uri_get_host (uri); +if (!uri_host ||g_str_equal (uri_host,"" )) {error_setg (errp,"missing hostname in URI" ); goto err; } -if (!uri->path ||strcmp (uri->path,"" ) ==0 ) { + uri_path =g_uri_get_path (uri); +if (!uri_path ||g_str_equal (uri_path,"" )) {error_setg (errp,"missing remote path in URI" ); goto err; } - qp =query_params_parse (uri->query); -if (!qp) { -error_setg (errp,"could not parse query parameters" ); - goto err; - } - -if (uri->user &&strcmp (uri->user,"" ) !=0 ) { -qdict_put_str (options,"user" , uri->user); + uri_user =g_uri_get_user (uri); +if (uri_user && !g_str_equal (uri_user,"" )) { +qdict_put_str (options,"user" , uri_user); } -qdict_put_str (options,"server.host" , uri->server); +qdict_put_str (options,"server.host" , uri_host); - port_str =g_strdup_printf ("%d" , uri->port ?:22 ); + port =g_uri_get_port (uri); + port_str =g_strdup_printf ("%d" , port !=-1 ? port :22 );qdict_put_str (options,"server.port" , port_str);g_free (port_str); -qdict_put_str (options,"path" , uri->path); - -/* Pick out any query parameters that we understand, and ignore -* the rest. -*/ -for (i =0 ; i < qp->n; ++i) { -if (strcmp (qp->p[i].name,"host_key_check" ) ==0 ) { -qdict_put_str (options,"host_key_check" , qp->p[i].value); +qdict_put_str (options,"path" , uri_path); + + uri_query =g_uri_get_query (uri); +if (uri_query) { +g_uri_params_iter_init (&qp, uri_query,-1 ,"&" ,G_URI_PARAMS_NONE ); +while (g_uri_params_iter_next (&qp, &qp_name, &qp_value, &gerror)) { +if (!qp_name || !qp_value || gerror) { +warn_report ("Failed to parse SSH URI parameters '%s'." , + uri_query); +break ; + } +/* +* Pick out the query parameters that we understand, and ignore +* (or rather warn about) the rest. +*/ +if (g_str_equal (qp_name,"host_key_check" )) { +qdict_put_str (options,"host_key_check" , qp_value); + }else { +warn_report ("Unsupported parameter '%s' in URI." , qp_name); + } } } -query_params_free (qp); -uri_free (uri); +g_uri_unref (uri);return 0 ; err: -uri_free (uri); +g_uri_unref (uri);return -EINVAL ; }
Signature · 1 line
2.44.0
Re: [PATCH for-9.1 6/9] block/nbd: Use URI parsing code from glib
In reply to Thomas HuthQuoted 133 lines · Thomas Huth
Since version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter. The
g_uri_get_host()also takes care of removing the square brackets from IPv6 addresses, so we can drop that part of the QEMU code now, too.Signed-off-by Thomas Huth <[email protected]>--- block/nbd.c | 66 ++++++++++++++++++++++++++++++----------------------- 1 file changed, 38 insertions(+), 28 deletions(-)block/nbd.c +36
-26@@ -31,7 +31,6 @@#include "qemu/osdep.h" #include "trace.h" -#include "qemu/uri.h" #include "qemu/option.h" #include "qemu/cutils.h" #include "qemu/main-loop.h" @@ -1514,30 +1513,34 @@ static void nbd_client_close(BlockDriverState *bs)static int nbd_parse_uri (const char *filename,QDict *options) { -URI *uri; +GUri *uri;const char *p; -QueryParams *qp =NULL ; +GHashTable *qp =NULL ; +int qp_n;int ret =0 ;bool is_unix; +const char *uri_scheme, *uri_query, *uri_server; +int uri_port; - uri =uri_parse (filename); + uri =g_uri_parse (filename,G_URI_FLAGS_NONE ,NULL );if (!uri) {return -EINVAL ; }/* transport */ -if (!g_strcmp0 (uri->scheme,"nbd" )) { + uri_scheme =g_uri_get_scheme (uri); +if (!g_strcmp0 (uri_scheme,"nbd" )) { is_unix = false; - }else if (!g_strcmp0 (uri->scheme,"nbd+tcp" )) { + }else if (!g_strcmp0 (uri_scheme,"nbd+tcp" )) { is_unix = false; - }else if (!g_strcmp0 (uri->scheme,"nbd+unix" )) { + }else if (!g_strcmp0 (uri_scheme,"nbd+unix" )) { is_unix = true; }else { ret = -EINVAL ; goto out; } - p = uri->path ? uri->path :"" ; + p =g_uri_get_path (uri) ?:"" ;if (p[0 ] =='/' ) { p++; }@@ -1545,51 +1548,58 @@ static int nbd_parse_uri(const char *filename, QDict *options)qdict_put_str (options,"export" , p); } - qp =query_params_parse (uri->query); -if (qp->n >1 || (is_unix && !qp->n) || (!is_unix && qp->n)) { - ret = -EINVAL ; - goto out; + uri_query =g_uri_get_query (uri); +if (uri_query) { + qp =g_uri_parse_params (uri_query,-1 ,"&" ,G_URI_PARAMS_NONE ,NULL ); +if (!qp) { + ret = -EINVAL ; + goto out; + } + qp_n =g_hash_table_size (qp); +if (qp_n >1 || (is_unix && !qp_n) || (!is_unix && qp_n)) { + ret = -EINVAL ; + goto out; + } + } + + uri_server =g_uri_get_host (uri); +if (uri_server && !uri_server[0 ]) { + uri_server =NULL ; } + uri_port =g_uri_get_port (uri);if (is_unix) {/* nbd+unix:///export?socket=path */ -if (uri->server || uri->port ||strcmp (qp->p[0 ].name,"socket" )) { +const char *uri_socket =g_hash_table_lookup (qp,"socket" ); +if (uri_server || uri_port !=-1 || !uri_socket) { ret = -EINVAL ; goto out; }qdict_put_str (options,"server.type" ,"unix" ); -qdict_put_str (options,"server.path" , qp->p[0 ].value); +qdict_put_str (options,"server.path" , uri_socket); }else { -QString *host;char *port_str;/* nbd[+tcp]://host[:port]/export */ -if (!uri->server) { +if (!uri_server) { ret = -EINVAL ; goto out; } -/* strip braces from literal IPv6 address */ -if (uri->server[0 ] =='[' ) { - host =qstring_from_substr (uri->server,1 , -strlen (uri->server) -1 ); - }else { - host =qstring_from_str (uri->server); - } -qdict_put_str (options,"server.type" ,"inet" ); -qdict_put (options,"server.host" , host); +qdict_put_str (options,"server.host" , uri_server); - port_str =g_strdup_printf ("%d" , uri->port ?:NBD_DEFAULT_PORT ); + port_str =g_strdup_printf ("%d" , uri_port !=-1 ? uri_port + :NBD_DEFAULT_PORT );qdict_put_str (options,"server.port" , port_str);g_free (port_str); }
block/nbd.c static int nbd_parse_uri(const char *filename, QDict *options) · line 1604
Quoted 10 lines
block/nbd.c +2
-2out:if (qp) { -query_params_free (qp); +g_hash_table_destroy (qp); } -uri_free (uri); +g_uri_unref (uri);return ret; }
Looks ok,
Signature · 4 lines
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://libguestfs.org
Re: [PATCH for-9.1 8/9] block/ssh: Use URI parsing code from glib
In reply to Thomas HuthQuoted 115 lines · Thomas Huth
Since version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter.
Signed-off-by Thomas Huth <[email protected]>--- block/ssh.c | 69 +++++++++++++++++++++++++++++++---------------------- 1 file changed, 40 insertions(+), 29 deletions(-)block/ssh.c +38
-26@@ -37,7 +37,6 @@#include "qemu/ctype.h" #include "qemu/cutils.h" #include "qemu/sockets.h" -#include "qemu/uri.h" #include "qapi/qapi-visit-sockets.h" #include "qapi/qapi-visit-block-core.h" #include "qapi/qmp/qdict.h" @@ -181,64 +180,76 @@ static void sftp_error_trace(BDRVSSHState *s, const char *op)static int parse_uri (const char *filename,QDict *options,Error **errp) { -URI *uri =NULL ; -QueryParams *qp; +GUri *uri; +const char *uri_host, *uri_path, *uri_user, *uri_query;char *port_str; -int i; +int port; +g_autoptr (GError) gerror =NULL ; +char *qp_name, *qp_value; +GUriParamsIter qp; - uri =uri_parse (filename); + uri =g_uri_parse (filename,G_URI_FLAGS_NONE ,NULL );if (!uri) {return -EINVAL ; } -if (g_strcmp0 (uri->scheme,"ssh" ) !=0 ) { +if (g_strcmp0 (g_uri_get_scheme (uri),"ssh" ) !=0 ) {error_setg (errp,"URI scheme must be 'ssh'" ); goto err; } -if (!uri->server ||strcmp (uri->server,"" ) ==0 ) { + uri_host =g_uri_get_host (uri); +if (!uri_host ||g_str_equal (uri_host,"" )) {error_setg (errp,"missing hostname in URI" ); goto err; } -if (!uri->path ||strcmp (uri->path,"" ) ==0 ) { + uri_path =g_uri_get_path (uri); +if (!uri_path ||g_str_equal (uri_path,"" )) {error_setg (errp,"missing remote path in URI" ); goto err; } - qp =query_params_parse (uri->query); -if (!qp) { -error_setg (errp,"could not parse query parameters" ); - goto err; - } - -if (uri->user &&strcmp (uri->user,"" ) !=0 ) { -qdict_put_str (options,"user" , uri->user); + uri_user =g_uri_get_user (uri); +if (uri_user && !g_str_equal (uri_user,"" )) { +qdict_put_str (options,"user" , uri_user); } -qdict_put_str (options,"server.host" , uri->server); +qdict_put_str (options,"server.host" , uri_host); - port_str =g_strdup_printf ("%d" , uri->port ?:22 ); + port =g_uri_get_port (uri); + port_str =g_strdup_printf ("%d" , port !=-1 ? port :22 );qdict_put_str (options,"server.port" , port_str);g_free (port_str); -qdict_put_str (options,"path" , uri->path); - -/* Pick out any query parameters that we understand, and ignore -* the rest. -*/ -for (i =0 ; i < qp->n; ++i) { -if (strcmp (qp->p[i].name,"host_key_check" ) ==0 ) { -qdict_put_str (options,"host_key_check" , qp->p[i].value); +qdict_put_str (options,"path" , uri_path); + + uri_query =g_uri_get_query (uri); +if (uri_query) { +g_uri_params_iter_init (&qp, uri_query,-1 ,"&" ,G_URI_PARAMS_NONE ); +while (g_uri_params_iter_next (&qp, &qp_name, &qp_value, &gerror)) { +if (!qp_name || !qp_value || gerror) { +warn_report ("Failed to parse SSH URI parameters '%s'." , + uri_query); +break ; + } +/* +* Pick out the query parameters that we understand, and ignore +* (or rather warn about) the rest. +*/ +if (g_str_equal (qp_name,"host_key_check" )) { +qdict_put_str (options,"host_key_check" , qp_value); + }else { +warn_report ("Unsupported parameter '%s' in URI." , qp_name); + } } }
block/ssh.c static void sftp_error_trace(BDRVSSHState *s, const char *op) · line 254
Quoted 10 lines
block/ssh.c +2
-3-query_params_free (qp); -uri_free (uri); +g_uri_unref (uri);return 0 ; err: -uri_free (uri); +g_uri_unref (uri);return -EINVAL ; }
Seems reasonable too,
Signature · 5 lines
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-p2v converts physical machines to virtual machines. Boot with a live CD or over the network (PXE) and turn machines into KVM guests. http://libguestfs.org/virt-v2v
Re: [PATCH for-9.1 6/9] block/nbd: Use URI parsing code from glib
In reply to Thomas HuthQuoted 24 lines · Thomas Huth
Since version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter. The
g_uri_get_host()also takes care of removing the square brackets from IPv6 addresses, so we can drop that part of the QEMU code now, too.Signed-off-by Thomas Huth <[email protected]>--- block/nbd.c | 66 ++++++++++++++++++++++++++++++----------------------- 1 file changed, 38 insertions(+), 28 deletions(-)block/nbd.c +0
-1@@ -31,7 +31,6 @@#include "qemu/osdep.h" #include "trace.h" -#include "qemu/uri.h" #include "qemu/option.h" #include "qemu/cutils.h" #include "qemu/main-loop.h"
block/nbd.c static void nbd_client_close(BlockDriverState *bs) · line 1516
Quoted 5 lines
block/nbd.c +1
-1@@ -1514,30 +1513,34 @@ static void nbd_client_close(BlockDriverState *bs)static int nbd_parse_uri (const char *filename,QDict *options) { -URI *uri; +GUri *uri;
Is it worth using 'g_autoptr(GUri) uri = NULL;' here, to simplify cleanup later?
block/nbd.c static void nbd_client_close(BlockDriverState *bs) · line 1518
Quoted 3 lines
block/nbd.c +1
-1const char *p; -QueryParams *qp =NULL ; +GHashTable *qp =NULL ;
Presumably would be easier if qp is also auto-free.
block/nbd.c static void nbd_client_close(BlockDriverState *bs) · line 1525
Quoted 8 lines
block/nbd.c +4
-1+int qp_n;int ret =0 ;bool is_unix; +const char *uri_scheme, *uri_query, *uri_server; +int uri_port; - uri =uri_parse (filename); + uri =g_uri_parse (filename,G_URI_FLAGS_NONE ,NULL );
The glib API is fairly close to what we have in qemu, making this a nice switchover.
Quoted 13 lines
block/nbd.c +1
-7/* nbd[+tcp]://host[:port]/export */ -if (!uri->server) { +if (!uri_server) { ret = -EINVAL ; goto out; } -/* strip braces from literal IPv6 address */ -if (uri->server[0 ] =='[' ) { - host =qstring_from_substr (uri->server,1 , -strlen (uri->server) -1 ); - }else { - host =qstring_from_str (uri->server);
block/nbd.c static int nbd_parse_uri(const char *filename, QDict *options) · line 1594
Quoted 10 lines
block/nbd.c +3
-4- } -qdict_put_str (options,"server.type" ,"inet" ); -qdict_put (options,"server.host" , host); +qdict_put_str (options,"server.host" , uri_server); - port_str =g_strdup_printf ("%d" , uri->port ?:NBD_DEFAULT_PORT ); + port_str =g_strdup_printf ("%d" , uri_port !=-1 ? uri_port + :NBD_DEFAULT_PORT );qdict_put_str (options,"server.port" , port_str);
If a user requests nbd://hostname:0/export, this now sets server.port to "0" instead of "10809". Is that an intentional change? No one actually passes an explicit ":0" port on purpose, but we do have to worry about malicious URIs.
block/nbd.c static int nbd_parse_uri(const char *filename, QDict *options) · line 1603
Quoted 11 lines
block/nbd.c +2
-2g_free (port_str); } out:if (qp) { -query_params_free (qp); +g_hash_table_destroy (qp); } -uri_free (uri); +g_uri_unref (uri);return ret;
It may be possible to eliminate the out label altogether, if GHashTable has the appropriate auto-free magic.
Signature · 3 lines
Eric Blake, Principal Software Engineer Red Hat, Inc. Virtualization: qemu.org | libguestfs.org
Re: [PATCH for-9.1 6/9] block/nbd: Use URI parsing code from glib
Quoted 75 lines · Eric Blake
Quoted 24 lines · Thomas Huth
Since version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter. The
g_uri_get_host()also takes care of removing the square brackets from IPv6 addresses, so we can drop that part of the QEMU code now, too.Signed-off-by Thomas Huth <[email protected]>--- block/nbd.c | 66 ++++++++++++++++++++++++++++++----------------------- 1 file changed, 38 insertions(+), 28 deletions(-)block/nbd.c +0
-1@@ -31,7 +31,6 @@#include "qemu/osdep.h" #include "trace.h" -#include "qemu/uri.h" #include "qemu/option.h" #include "qemu/cutils.h" #include "qemu/main-loop.h" block/nbd.c static void nbd_client_close(BlockDriverState *bs) · line 1516
Quoted 5 lines
block/nbd.c +1
-1@@ -1514,30 +1513,34 @@ static void nbd_client_close(BlockDriverState *bs)static int nbd_parse_uri (const char *filename,QDict *options) { -URI *uri; +GUri *uri;Is it worth using '
g_autoptr(GUri)uri = NULL;' here, to simplify cleanup later?block/nbd.c static void nbd_client_close(BlockDriverState *bs) · line 1518
Quoted 3 lines
block/nbd.c +1
-1const char *p; -QueryParams *qp =NULL ; +GHashTable *qp =NULL ;Presumably would be easier if qp is also auto-free.
block/nbd.c static void nbd_client_close(BlockDriverState *bs) · line 1525
Quoted 8 lines
block/nbd.c +4
-1+int qp_n;int ret =0 ;bool is_unix; +const char *uri_scheme, *uri_query, *uri_server; +int uri_port; - uri =uri_parse (filename); + uri =g_uri_parse (filename,G_URI_FLAGS_NONE ,NULL );The glib API is fairly close to what we have in qemu, making this a nice switchover.
Quoted 13 lines
block/nbd.c +1
-7/* nbd[+tcp]://host[:port]/export */ -if (!uri->server) { +if (!uri_server) { ret = -EINVAL ; goto out; } -/* strip braces from literal IPv6 address */ -if (uri->server[0 ] =='[' ) { - host =qstring_from_substr (uri->server,1 , -strlen (uri->server) -1 ); - }else { - host =qstring_from_str (uri->server);block/nbd.c static int nbd_parse_uri(const char *filename, QDict *options) · line 1594
Quoted 10 lines
block/nbd.c +3
-4- } -qdict_put_str (options,"server.type" ,"inet" ); -qdict_put (options,"server.host" , host); +qdict_put_str (options,"server.host" , uri_server); - port_str =g_strdup_printf ("%d" , uri->port ?:NBD_DEFAULT_PORT ); + port_str =g_strdup_printf ("%d" , uri_port !=-1 ? uri_port + :NBD_DEFAULT_PORT );qdict_put_str (options,"server.port" , port_str);
Quoted 4 lines
If a user requests nbd://hostname:0/export, this now sets server.port to "0" instead of "10809". Is that an intentional change? No one actually passes an explicit ":0" port on purpose, but we do have to worry about malicious URIs.
Passing '0' will cause the kernel to allocate a random free port, so that is potentially introducing new semantics ?
With regards,
Daniel
Signature · 3 lines
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
Re: [PATCH for-9.1 6/9] block/nbd: Use URI parsing code from glib
In reply to Richard W.M. JonesAdjusting cc list to add upstream NBD and drop developers unrelated to this part of the qemu series...
Quoted 9 lines · Richard W.M. Jones
Quoted 8 lines · Thomas Huth
Since version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter. The
g_uri_get_host()also takes care of removing the square brackets from IPv6 addresses, so we can drop that part of the QEMU code now, too.
Quoted 8 lines
block/nbd.c static int nbd_parse_uri(const char *filename, QDict *options) · line 1577
Quoted 8 lines
if (is_unix) { /* nbd+unix:///export?socket=path */block/nbd.c +2
-1-if (uri->server || uri->port ||strcmp (qp->p[0 ].name,"socket" )) { +const char *uri_socket =g_hash_table_lookup (qp,"socket" ); +if (uri_server || uri_port !=-1 || !uri_socket) { ret = -EINVAL ; goto out; }
The spec for NBD URIs is at:
https://github.com/NetworkBlockDevice/nbd/blob/master/doc/uri.md
Should any of this spec mention case-insensitive concerns, such as whether 'NBD://' may be equivalent to 'nbd://', and whether 'nbd+unix:///?socket=x' is equivalent to 'nbd+unix:///?Socket=x'? Right now, I think that most implementations of NBD servers and clients happen to use case-sensitive parsing; but glib provides the option to do case-insensitive query parsing.
If I read https://docs.gtk.org/glib/type_func.Uri.parse_params.html
correctly, passing G_URI_PARAMS_CASE_INSENSITIVE (which you did not do) would mean that 'nbd+unix:///?socket=ignore&Socket=/for/real' would result in this g_hash_table_lookup finding only "Socket", not "socket". Maybe it is worth an explicit addition to the NBD URI spec to mention that we intend to be case-sensitive (in the parts where it can be; I'm not sure if the schema part must be handled case-insensitively without re-reading the RFCs), and therefore that 'Socket=' does NOT have the same meaning as 'socket='.
Signature · 3 lines
Eric Blake, Principal Software Engineer Red Hat, Inc. Virtualization: qemu.org | libguestfs.org
Re: [PATCH for-9.1 6/9] block/nbd: Use URI parsing code from glib
In reply to Daniel P. BerrangéQuoted 27 lines · Daniel P. Berrangé
Quoted 26 lines · Eric Blake
Quoted 25 lines · Thomas Huth
Since version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter. The
g_uri_get_host()also takes care of removing the square brackets from IPv6 addresses, so we can drop that part of the QEMU code now, too.Signed-off-by Thomas Huth <[email protected]>--- block/nbd.c | 66 ++++++++++++++++++++++++++++++----------------------- 1 file changed, 38 insertions(+), 28 deletions(-)block/nbd.c +0
-1@@ -31,7 +31,6 @@#include "qemu/osdep.h" #include "trace.h" -#include "qemu/uri.h" #include "qemu/option.h" #include "qemu/cutils.h" #include "qemu/main-loop.h" @@ -1514,30 +1513,34 @@ static void nbd_client_close(BlockDriverState *bs)
Quoted 6 lines
Quoted 6 lines
block/nbd.c static void nbd_client_close(BlockDriverState *bs) · line 1516
Quoted 4 lines
block/nbd.c +1
-1static int nbd_parse_uri (const char *filename,QDict *options) { -URI *uri; +GUri *uri;Is it worth using '
g_autoptr(GUri)uri = NULL;' here, to simplify cleanup later?
Sounds like a good idea, I'll give it a try!
Quoted 42 lines
Quoted 42 lines
block/nbd.c static void nbd_client_close(BlockDriverState *bs) · line 1518
Quoted 3 lines
block/nbd.c +1
-1const char *p; -QueryParams *qp =NULL ; +GHashTable *qp =NULL ;Presumably would be easier if qp is also auto-free.
block/nbd.c static void nbd_client_close(BlockDriverState *bs) · line 1525
Quoted 8 lines
block/nbd.c +4
-1+int qp_n;int ret =0 ;bool is_unix; +const char *uri_scheme, *uri_query, *uri_server; +int uri_port; - uri =uri_parse (filename); + uri =g_uri_parse (filename,G_URI_FLAGS_NONE ,NULL );The glib API is fairly close to what we have in qemu, making this a nice switchover.
Quoted 23 lines
block/nbd.c +4
-11/* nbd[+tcp]://host[:port]/export */ -if (!uri->server) { +if (!uri_server) { ret = -EINVAL ; goto out; } -/* strip braces from literal IPv6 address */ -if (uri->server[0 ] =='[' ) { - host =qstring_from_substr (uri->server,1 , -strlen (uri->server) -1 ); - }else { - host =qstring_from_str (uri->server); - } -qdict_put_str (options,"server.type" ,"inet" ); -qdict_put (options,"server.host" , host); +qdict_put_str (options,"server.host" , uri_server); - port_str =g_strdup_printf ("%d" , uri->port ?:NBD_DEFAULT_PORT ); + port_str =g_strdup_printf ("%d" , uri_port !=-1 ? uri_port + :NBD_DEFAULT_PORT );qdict_put_str (options,"server.port" , port_str);
Quoted 7 lines
Quoted 4 lines
If a user requests nbd://hostname:0/export, this now sets server.port to "0" instead of "10809". Is that an intentional change? No one actually passes an explicit ":0" port on purpose, but we do have to worry about malicious URIs.
Passing '0' will cause the kernel to allocate a random free port, so that is potentially introducing new semantics ?
Ok, so passing a 0 does not really make sense, since QEMU needs to know the exact port. I'll change the check from "uri_port != -1" to "uri_port > 0" in the next version.
Thomas
Re: [PATCH for-9.1 6/9] block/nbd: Use URI parsing code from glib
In reply to Eric BlakeQuoted 24 lines · Eric Blake
Adjusting cc list to add upstream NBD and drop developers unrelated to this part of the qemu series...
Quoted 9 lines · Richard W.M. Jones
Quoted 8 lines · Thomas Huth
Since version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter. The
g_uri_get_host()also takes care of removing the square brackets from IPv6 addresses, so we can drop that part of the QEMU code now, too.Quoted 8 lines
block/nbd.c static int nbd_parse_uri(const char *filename, QDict *options) · line 1577
Quoted 8 lines
if (is_unix) { /* nbd+unix:///export?socket=path */block/nbd.c +2
-1-if (uri->server || uri->port ||strcmp (qp->p[0 ].name,"socket" )) { +const char *uri_socket =g_hash_table_lookup (qp,"socket" ); +if (uri_server || uri_port !=-1 || !uri_socket) { ret = -EINVAL ; goto out; }The spec for NBD URIs is at:
Quoted 8 lines
https://github.com/NetworkBlockDevice/nbd/blob/master/doc/uri.md
Should any of this spec mention case-insensitive concerns, such as whether 'NBD://' may be equivalent to 'nbd://', and whether 'nbd+unix:///?socket=x' is equivalent to 'nbd+unix:///?Socket=x'? Right now, I think that most implementations of NBD servers and clients happen to use case-sensitive parsing; but glib provides the option to do case-insensitive query parsing.
I haven't thought about this before, but do note that the NBD URI spec defers to "IETF standards describing URIs" for all unanswered questions. RFC3986 does talk about this incidentally. About the scheme field it says (section 3.1):
Although schemes are case-
insensitive, the canonical form is lowercase and documents that specify schemes must do so with lowercase letters. An implementation should accept uppercase letters as equivalent to lowercase in scheme names (e.g., allow "HTTP" as well as "http") for the sake of robustness but should only produce lowercase scheme names for consistency.
Hostname is also (obviously) case insensitive. There's also a section (6.2.3) which talks about normalization of URIs.
Overall it seems the intention of the RFC writer is that parsers should handle any case; but when generating URIs (and for examples, documentation etc) we should only generate lowercase.
libnbd absolutely does *not* get this right, eg:
$ nbdinfo NBD://localhost nbdinfo: nbd_connect_uri: unknown NBD URI scheme: NBD: Invalid argument
so that's a bug too.
Quoted 9 lines
If I read https://docs.gtk.org/glib/type_func.Uri.parse_params.html
correctly, passingG_URI_PARAMS_CASE_INSENSITIVE(which you did not do) would mean that 'nbd+unix:///?socket=ignore&Socket=/for/real' would result in thisg_hash_table_lookupfinding only "Socket", not "socket". Maybe it is worth an explicit addition to the NBD URI spec to mention that we intend to be case-sensitive (in the parts where it can be; I'm not sure if the schema part must be handled case-insensitively without re-reading the RFCs), and therefore that 'Socket=' does NOT have the same meaning as 'socket='.
We could mention this in the spec for clarity, but the current meaning (as above) would be that case-insensitive parsing is recommended.
Rich.
Signature · 4 lines
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://libguestfs.org
Re: [PATCH for-9.1 6/9] block/nbd: Use URI parsing code from glib
Quoted 6 lines · Richard W.M. Jones
libnbd absolutely does *not* get this right, eg:
$ nbdinfo NBD://localhost nbdinfo: nbd_connect_uri: unknown NBD URI scheme: NBD: Invalid argumentso that's a bug too.
Proposed fix:
https://gitlab.com/nbdkit/libnbd/-/merge_requests/6
Rich.
Signature · 5 lines
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://people.redhat.com/~rjones/virt-top
Re: [PATCH for-9.1 4/9] Bump minimum glib version to v2.66
In reply to Thomas HuthQuoted 15 lines · Thomas Huth
Now that we dropped support for CentOS 8 and Ubuntu 20.04, we can look into bumping the glib version to a new minimum for further clean-ups. According to repology.org, available versions are:
CentOS Stream 9: 2.66.7 Debian 11: 2.66.8 Fedora 38: 2.74.1 Freebsd: 2.78.4 Homebrew: 2.80.0 Openbsd: 2.78.4 OpenSuse leap 15.5: 2.70.5 pkgsrc_current: 2.78.4 Ubuntu 22.04: 2.72.1
Quoted 3 lines
Thus it should be safe to bump the minimum glib version to 2.66 now. Version 2.66 comes with new functions for URI parsing which will allow further clean-ups in the following patches.
Missing:
qga/commands-posix-ssh.c +0 -8
@@ -288,7 +288,6 @@ qmp_guest_ssh_get_authorized_keys( }#ifdef QGA_BUILD_UNIT_TEST -#if GLIB_CHECK_VERSION (2 ,60 ,0 )static const strList test_key2 = { .value = (char *)"algo key2 comments" };
@@ -484,11 +483,4 @@ int main(int argc, char *argv[]) return g_test_run(); } -#else -int main (int argc,char *argv[]) -{ -g_test_message ("test skipped, needs glib >= 2.60" ); -return 0 ; -} -#endif /* GLIB_2_60 */ #endif /* BUILD_UNIT_TEST */
Paolo
Re: [PATCH for-9.1 4/9] Bump minimum glib version to v2.66
Quoted 36 lines · Paolo Bonzini
Quoted 15 lines · Thomas Huth
Now that we dropped support for CentOS 8 and Ubuntu 20.04, we can look into bumping the glib version to a new minimum for further clean-ups. According to repology.org, available versions are:
CentOS Stream 9: 2.66.7 Debian 11: 2.66.8 Fedora 38: 2.74.1 Freebsd: 2.78.4 Homebrew: 2.80.0 Openbsd: 2.78.4 OpenSuse leap 15.5: 2.70.5 pkgsrc_current: 2.78.4 Ubuntu 22.04: 2.72.1Quoted 3 lines
Thus it should be safe to bump the minimum glib version to 2.66 now. Version 2.66 comes with new functions for URI parsing which will allow further clean-ups in the following patches.
Missing:
qga/commands-posix-ssh.c +0
-1@@ -288,7 +288,6 @@ qmp_guest_ssh_get_authorized_keys( }#ifdef QGA_BUILD_UNIT_TEST -#if GLIB_CHECK_VERSION (2 ,60 ,0 )static const strList test_key2 = { .value = (char *)"algo key2 comments" };@@ -484,11 +483,4 @@ int main(int argc, char *argv[])
qga/commands-posix-ssh.c int main(int argc, char *argv[]) · line 486
Quoted 10 lines
qga/commands-posix-ssh.c +0
-7return g_test_run(); } -#else -int main (int argc,char *argv[]) -{ -g_test_message ("test skipped, needs glib >= 2.60" ); -return 0 ; -} -#endif /* GLIB_2_60 */ #endif /* BUILD_UNIT_TEST */
Indeed! And there seems to be another GLIB_CHECK_VERSION(2,62,0) check in util/error-report.c which we likely can clean up now, too!
Thomas
Re: [PATCH for-9.1 4/9] Bump minimum glib version to v2.66
Quoted 37 lines · Thomas Huth
Quoted 36 lines · Paolo Bonzini
Quoted 15 lines · Thomas Huth
Now that we dropped support for CentOS 8 and Ubuntu 20.04, we can look into bumping the glib version to a new minimum for further clean-ups. According to repology.org, available versions are:
CentOS Stream 9: 2.66.7
Debian 11: 2.66.8
Fedora 38: 2.74.1
Freebsd: 2.78.4
Homebrew: 2.80.0
Openbsd: 2.78.4
OpenSuse leap 15.5: 2.70.5
pkgsrc_current: 2.78.4
Ubuntu 22.04: 2.72.1Quoted 3 lines
Thus it should be safe to bump the minimum glib version to 2.66 now. Version 2.66 comes with new functions for URI parsing which will allow further clean-ups in the following patches.
Missing:
qga/commands-posix-ssh.c +0
-1@@ -288,7 +288,6 @@ qmp_guest_ssh_get_authorized_keys( } #ifdef QGA_BUILD_UNIT_TEST -#if GLIB_CHECK_VERSION (2 ,60 ,0 ) static const strList test_key2 = {.value = (char *)"algo key2 comments"
};qga/commands-posix-ssh.c +0
-0@@ -484,11 +483,4 @@ int main(int argc, char *argv[])
Quoted 12 lines
qga/commands-posix-ssh.c int main(int argc, char *argv[]) · line 486
Quoted 10 lines
qga/commands-posix-ssh.c +0
-7return g_test_run(); } -#else -int main (int argc,char *argv[]) -{ - g_test_message("test skipped, needs glib >= 2.60"); - return 0; -} -#endif /* GLIB_2_60 */ #endif /* BUILD_UNIT_TEST */Indeed! And there seems to be another
GLIB_CHECK_VERSION(2,62,0)check inutil/error-report.cwhich we likely can clean up now, too!
Ok, I'll squash the above and
util/error-report.c +0 -10
@@ -172,18 +172,8 @@ static void print_loc(void)static char *real_time_iso8601 (void ) { -#if GLIB_CHECK_VERSION (2 ,62 ,0 )g_autoptr (GDateTime) dt =g_date_time_new_now_utc (); -/* ignore deprecation warning, since GLIB_VERSION_MAX_ALLOWED is 2.56 */ -#pragma GCC diagnostic push -#pragma GCC diagnostic ignored "-Wdeprecated-declarations"return g_date_time_format_iso8601 (dt); -#pragma GCC diagnostic pop -#else -GTimeVal tv; -g_get_current_time (&tv); -return g_time_val_to_iso8601 (&tv); -#endif }/*
then. As an aside, we probably can also drop:
/*
* gtk_widget_set_double_buffered() was deprecated in 3.14.
* It is required for opengl rendering on X11 though. A
* proper replacement (native opengl support) is only
* available in 3.16+. Silence the warning if possible. */
#pragma GCC diagnostic push
#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
gtk_widget_set_double_buffered(vc->gfx.drawing_area, FALSE);
#pragma GCC diagnostic pop
and
#pragma GCC diagnostic push
#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
/*
* check if RBD image is a clone (= has a parent).
*
* rbd_get_parent_info is deprecated from Nautilus onwards, but the
* replacement rbd_get_parent is not present in Luminous and Mimic.
*/
if (rbd_get_parent_info(s->image, NULL, 0, NULL, 0, NULL, 0) != -ENOENT) {
return status;
}#pragma GCC diagnostic pop
(Nautilus is Ceph 14, it's in all of CentOS Stream 9, Ubuntu 20.04 and Debian 11) but I have no idea what the replacement would be. :/
Paolo
MAINTAINERS tweak [was: [PATCH for-9.1 0/9] Switch to glib URI parsing code]
In reply to Thomas Huth[Trying Peter Lieven's alternate address...]
Quoted 3 lines · Thomas Huth
In the QEMU 9.1 development cycle, we can drop the support for Ubuntu 20.04 and CentOS 8 since the following major versions of these distributions are available since 2 years already.
Every time I've replied to any message in this thread, I've gotten a response:
| +Your message to [email protected] couldn't be delivered. | kamp.de couldn't confirm that your message was sent from a trusted location. | eblake Office 365 pl | Action Required Recipient | SPF validation error | | How to Fix It | Your organization's email admin will have to diagnose and fix your domain's email settings. Please forward this message to your | +email admin. | | ________________________________ | | More Info for Email Admins | Status code: 550 5.7.23 | | This error occurs when Sender Policy Framework (SPF) validation for the sender's domain fails. If you're the sender's email | +admin, make sure the SPF records for your domain at your domain registrar are set up correctly. Office 365 supports only one | +SPF record (a TXT record that defines SPF) for your domain. Include the following domain name: spf.protection.outlook.com. If | +you have a hybrid configuration (some mailboxes in the cloud, and some mailboxes on premises) or if you're an Exchange Online | +Protection standalone customer, add the outbound IP address of your on-premises servers to the TXT record.
Red Hat IT says that it is unlikely to be Red Hat's SPF settings, and suspects that it is instead something caused by whatever Peter is using to bounce mail from his alias Peter Lieven <[email protected]> to his Office 365 account. As I appear to be unable to contact Peter (even my use of direct email, bypassing the list, and using a personal account instead of my Red Hat email) about this issue, I'm wondering if Peter is still an active contributor to the project.
But while typing this email, to see if RBD, iSCSI, and NFS need a new entry in MAINTAINERS, I did a search through the list archives, where the last email I found from Peter was
https://lists.gnu.org/archive/html/qemu-devel/2023-01/msg00574.html,
which asked to update MAINTAINERS to his new address, and that has not made it in so far...
Signature · 3 lines
Eric Blake, Principal Software Engineer Red Hat, Inc. Virtualization: qemu.org | libguestfs.org
Re: [PATCH for-9.1 6/9] block/nbd: Use URI parsing code from glib
In reply to Richard W.M. JonesQuoted 147 lines · Richard W.M. Jones
Quoted 133 lines · Thomas Huth
Since version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter. The
g_uri_get_host()also takes care of removing the square brackets from IPv6 addresses, so we can drop that part of the QEMU code now, too.Signed-off-by Thomas Huth <[email protected]>--- block/nbd.c | 66 ++++++++++++++++++++++++++++++----------------------- 1 file changed, 38 insertions(+), 28 deletions(-)block/nbd.c +36
-26@@ -31,7 +31,6 @@#include "qemu/osdep.h" #include "trace.h" -#include "qemu/uri.h" #include "qemu/option.h" #include "qemu/cutils.h" #include "qemu/main-loop.h" @@ -1514,30 +1513,34 @@ static void nbd_client_close(BlockDriverState *bs)static int nbd_parse_uri (const char *filename,QDict *options) { -URI *uri; +GUri *uri;const char *p; -QueryParams *qp =NULL ; +GHashTable *qp =NULL ; +int qp_n;int ret =0 ;bool is_unix; +const char *uri_scheme, *uri_query, *uri_server; +int uri_port; - uri =uri_parse (filename); + uri =g_uri_parse (filename,G_URI_FLAGS_NONE ,NULL );if (!uri) {return -EINVAL ; }/* transport */ -if (!g_strcmp0 (uri->scheme,"nbd" )) { + uri_scheme =g_uri_get_scheme (uri); +if (!g_strcmp0 (uri_scheme,"nbd" )) { is_unix = false; - }else if (!g_strcmp0 (uri->scheme,"nbd+tcp" )) { + }else if (!g_strcmp0 (uri_scheme,"nbd+tcp" )) { is_unix = false; - }else if (!g_strcmp0 (uri->scheme,"nbd+unix" )) { + }else if (!g_strcmp0 (uri_scheme,"nbd+unix" )) { is_unix = true; }else { ret = -EINVAL ; goto out; } - p = uri->path ? uri->path :"" ; + p =g_uri_get_path (uri) ?:"" ;if (p[0 ] =='/' ) { p++; }@@ -1545,51 +1548,58 @@ static int nbd_parse_uri(const char *filename, QDict *options)qdict_put_str (options,"export" , p); } - qp =query_params_parse (uri->query); -if (qp->n >1 || (is_unix && !qp->n) || (!is_unix && qp->n)) { - ret = -EINVAL ; - goto out; + uri_query =g_uri_get_query (uri); +if (uri_query) { + qp =g_uri_parse_params (uri_query,-1 ,"&" ,G_URI_PARAMS_NONE ,NULL ); +if (!qp) { + ret = -EINVAL ; + goto out; + } + qp_n =g_hash_table_size (qp); +if (qp_n >1 || (is_unix && !qp_n) || (!is_unix && qp_n)) { + ret = -EINVAL ; + goto out; + } + } + + uri_server =g_uri_get_host (uri); +if (uri_server && !uri_server[0 ]) { + uri_server =NULL ; } + uri_port =g_uri_get_port (uri);if (is_unix) {/* nbd+unix:///export?socket=path */ -if (uri->server || uri->port ||strcmp (qp->p[0 ].name,"socket" )) { +const char *uri_socket =g_hash_table_lookup (qp,"socket" ); +if (uri_server || uri_port !=-1 || !uri_socket) { ret = -EINVAL ; goto out; }qdict_put_str (options,"server.type" ,"unix" ); -qdict_put_str (options,"server.path" , qp->p[0 ].value); +qdict_put_str (options,"server.path" , uri_socket); }else { -QString *host;char *port_str;/* nbd[+tcp]://host[:port]/export */ -if (!uri->server) { +if (!uri_server) { ret = -EINVAL ; goto out; } -/* strip braces from literal IPv6 address */ -if (uri->server[0 ] =='[' ) { - host =qstring_from_substr (uri->server,1 , -strlen (uri->server) -1 ); - }else { - host =qstring_from_str (uri->server); - } -qdict_put_str (options,"server.type" ,"inet" ); -qdict_put (options,"server.host" , host); +qdict_put_str (options,"server.host" , uri_server); - port_str =g_strdup_printf ("%d" , uri->port ?:NBD_DEFAULT_PORT ); + port_str =g_strdup_printf ("%d" , uri_port !=-1 ? uri_port + :NBD_DEFAULT_PORT );qdict_put_str (options,"server.port" , port_str);g_free (port_str); }block/nbd.c static int nbd_parse_uri(const char *filename, QDict *options) · line 1604
Quoted 10 lines
block/nbd.c +2
-2out:if (qp) { -query_params_free (qp); +g_hash_table_destroy (qp); } -uri_free (uri); +g_uri_unref (uri);return ret; }Looks ok,
Quoted 1 line
Reviewed-by Richard W.M. Jones <[email protected]>
Or maybe not. This caused a regression in the nbdkit test suite (when we use qemu-img from 9.1). It seems the exportname part of the NBD URI gets munged:
https://gitlab.com/qemu-project/qemu/-/issues/2584
Rich.
Signature · 4 lines
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-builder quickly builds VMs from scratch http://libguestfs.org/virt-builder.1.html
Re: [PATCH for-9.1 6/9] block/nbd: Use URI parsing code from glib
Quoted 18 lines · Richard W.M. Jones
Quoted 17 lines · Richard W.M. Jones
block/nbd.c static void nbd_client_close(BlockDriverState *bs) · line 1546
Quoted 13 lines · Thomas Huth
Since version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter. The
g_uri_get_host()also takes care of removing the square brackets from IPv6 addresses, so we can drop that part of the QEMU code now, too.block/nbd.c +1
-1- p = uri->path ? uri->path :"" ; + p =g_uri_get_path (uri) ?:"" ;if (p[0 ] =='/' ) { p++; }Looks ok,
Quoted 7 lines
Quoted 1 line
Reviewed-by Richard W.M. Jones <[email protected]>Or maybe not. This caused a regression in the nbdkit test suite (when we use qemu-img from 9.1). It seems the exportname part of the NBD URI gets munged:
To be more specific, it looks like g_uri_get_path("./name//with//..//slashes") is getting munged to "name/slashes". That is, glib is blindly assuming that ./ and XXX/../ can be dropped, and // can be simplified to /, which may be true for arbitrary file names but not true for abitrary URIs (since URIs have application-specific semantics, which may not match path name traversal semantics). Looks like we need to report a bug to glib, and/or see if glib's URI functions have a flag for turning off this unwanted munging.
Or we may just want to document this corner case change as intentional.
Signature · 3 lines
Eric Blake, Principal Software Engineer Red Hat, Inc. Virtualization: qemu.org | libguestfs.org
Re: [PATCH for-9.1 6/9] block/nbd: Use URI parsing code from glib
Quoted 27 lines · Eric Blake
Quoted 18 lines · Richard W.M. Jones
Quoted 17 lines · Richard W.M. Jones
block/nbd.c static void nbd_client_close(BlockDriverState *bs) · line 1546
Quoted 13 lines · Thomas Huth
Since version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter. The
g_uri_get_host()also takes care of removing the square brackets from IPv6 addresses, so we can drop that part of the QEMU code now, too.block/nbd.c +1
-1- p = uri->path ? uri->path :"" ; + p =g_uri_get_path (uri) ?:"" ;if (p[0 ] =='/' ) { p++; }Looks ok,
Quoted 7 lines
Quoted 1 line
Reviewed-by Richard W.M. Jones <[email protected]>Or maybe not. This caused a regression in the nbdkit test suite (when we use qemu-img from 9.1). It seems the exportname part of the NBD URI gets munged:
Quoted 9 lines
To be more specific, it looks like
g_uri_get_path("./name//with//..//slashes")is getting munged to "name/slashes". That is, glib is blindly assuming that ./ and XXX/../ can be dropped, and // can be simplified to /, which may be true for arbitrary file names but not true for abitrary URIs (since URIs have application-specific semantics, which may not match path name traversal semantics). Looks like we need to report a bug to glib, and/or see if glib's URI functions have a flag for turning off this unwanted munging.
The source code indicates it is doing some normalization based on this:
https://datatracker.ietf.org/doc/html/rfc3986#section-5.2.4
With regards,
Daniel
Signature · 3 lines
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
Re: [PATCH for-9.1 6/9] block/nbd: Use URI parsing code from glib
Quoted 38 lines · Daniel P. Berrangé
Quoted 27 lines · Eric Blake
Quoted 18 lines · Richard W.M. Jones
Quoted 17 lines · Richard W.M. Jones
block/nbd.c static void nbd_client_close(BlockDriverState *bs) · line 1546
Quoted 13 lines · Thomas Huth
Since version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter. The
g_uri_get_host()also takes care of removing the square brackets from IPv6 addresses, so we can drop that part of the QEMU code now, too.block/nbd.c +1
-1- p = uri->path ? uri->path :"" ; + p =g_uri_get_path (uri) ?:"" ;if (p[0 ] =='/' ) { p++; }Looks ok,
Quoted 7 lines
Quoted 1 line
Reviewed-by Richard W.M. Jones <[email protected]>Or maybe not. This caused a regression in the nbdkit test suite (when we use qemu-img from 9.1). It seems the exportname part of the NBD URI gets munged:
Quoted 9 lines
To be more specific, it looks like
g_uri_get_path("./name//with//..//slashes")is getting munged to "name/slashes". That is, glib is blindly assuming that ./ and XXX/../ can be dropped, and // can be simplified to /, which may be true for arbitrary file names but not true for abitrary URIs (since URIs have application-specific semantics, which may not match path name traversal semantics). Looks like we need to report a bug to glib, and/or see if glib's URI functions have a flag for turning off this unwanted munging.
Quoted 4 lines
The source code indicates it is doing some normalization based on this:
I wrote a bit about this in the bug:
https://gitlab.com/qemu-project/qemu/-/issues/2584#note_2125192404
Rich.
Signature · 4 lines
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://libguestfs.org
Re: [PATCH for-9.1 6/9] block/nbd: Use URI parsing code from glib
In reply to Eric BlakeQuoted 37 lines · Eric Blake
Quoted 18 lines · Richard W.M. Jones
Quoted 17 lines · Richard W.M. Jones
block/nbd.c static void nbd_client_close(BlockDriverState *bs) · line 1546
Quoted 13 lines · Thomas Huth
Since version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter. The
g_uri_get_host()also takes care of removing the square brackets from IPv6 addresses, so we can drop that part of the QEMU code now, too.block/nbd.c +1
-1- p = uri->path ? uri->path :"" ; + p =g_uri_get_path (uri) ?:"" ;if (p[0 ] =='/' ) { p++; }Looks ok,
Quoted 7 lines
Quoted 1 line
Reviewed-by Richard W.M. Jones <[email protected]>Or maybe not. This caused a regression in the nbdkit test suite (when we use qemu-img from 9.1). It seems the exportname part of the NBD URI gets munged:
To be more specific, it looks like
g_uri_get_path("./name//with//..//slashes")is getting munged to "name/slashes". That is, glib is blindly assuming that ./ and XXX/../ can be dropped, and // can be simplified to /, which may be true for arbitrary file names but not true for abitrary URIs (since URIs have application-specific semantics, which may not match path name traversal semantics). Looks like we need to report a bug to glib, and/or see if glib's URI functions have a flag for turning off this unwanted munging.
Quoted 2 lines
Or we may just want to document this corner case change as intentional.
Ok ... so how bad is this for NBD? Can we go along with the odditiy or is this breaking some real world NBD scenarios?
... in the worst case, we have to revert the patch ...
Thomas
Re: [PATCH for-9.1 6/9] block/nbd: Use URI parsing code from glib
Quoted 37 lines · Thomas Huth
Quoted 36 lines · Eric Blake
Quoted 17 lines · Richard W.M. Jones
Quoted 16 lines · Richard W.M. Jones
block/nbd.c static void nbd_client_close(BlockDriverState *bs) · line 1546
Quoted 12 lines · Thomas Huth
Since version 2.66, glib has useful URI parsing functions, too. Use those instead of the QEMU-internal ones to be finally able to get rid of the latter. The
g_uri_get_host()also takes care of removing the square brackets from IPv6 addresses, so we can drop that part of the QEMU code now, too.block/nbd.c +1
-1- p = uri->path ? uri->path :"" ; + p =g_uri_get_path (uri) ?:"" ;if (p[0 ] =='/' ) { p++; }Looks ok,
Quoted 7 lines
Quoted 1 line
Reviewed-by Richard W.M. Jones <[email protected]>Or maybe not. This caused a regression in the nbdkit test suite (when we use qemu-img from 9.1). It seems the exportname part of the NBD URI gets munged:
To be more specific, it looks like
g_uri_get_path("./name//with//..//slashes")is getting munged to "name/slashes". That is, glib is blindly assuming that ./ and XXX/../ can be dropped, and // can be simplified to /, which may be true for arbitrary file names but not true for abitrary URIs (since URIs have application-specific semantics, which may not match path name traversal semantics). Looks like we need to report a bug to glib, and/or see if glib's URI functions have a flag for turning off this unwanted munging.
Quoted 5 lines
Quoted 2 lines
Or we may just want to document this corner case change as intentional.
Ok ... so how bad is this for NBD? Can we go along with the odditiy or is this breaking some real world NBD scenarios?
After sleeping on it, I don't think it's serious at all. It's also permitted (albeit not required) by the URI RFCs, so other implementations in future might do the same thing.
Quoted 1 line
... in the worst case, we have to revert the patch ...
I closed the bug just now and have decided to go with a documentation update instead.
Rich.
Signature · 5 lines
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://people.redhat.com/~rjones/virt-top