March 2024 - Devel - libvirt List Archives

[PULL 30/43] hw/core/machine-smp: Remove deprecated "parameter=0" SMP configurations
by Philippe Mathieu-Daudé 09 Mar '24

09 Mar '24

From: Zhao Liu <zhao1.liu(a)intel.com> The "parameter=0" SMP configurations have been marked as deprecated since v6.2. For these cases, -smp currently returns the warning and adjusts the zeroed parameters to 1 by default. Remove the above compatibility logic in v9.0, and return error directly if any -smp parameter is set as 0. Signed-off-by: Zhao Liu <zhao1.liu(a)intel.com> Reviewed-by: Thomas Huth <thuth(a)redhat.com> Reviewed-by: Prasad Pandit <pjp(a)fedoraproject.org> Message-ID: <20240308160148.3130837-2-zhao1.liu(a)linux.intel.com> Signed-off-by: Philippe Mathieu-Daudé <philmd(a)linaro.org> --- docs/about/deprecated.rst | 16 ---------------- docs/about/removed-features.rst | 15 +++++++++++++++ hw/core/machine-smp.c | 5 +++-- 3 files changed, 18 insertions(+), 18 deletions(-) diff --git a/docs/about/deprecated.rst b/docs/about/deprecated.rst index 8565644da6..6e2f557682 100644 --- a/docs/about/deprecated.rst +++ b/docs/about/deprecated.rst @@ -36,22 +36,6 @@ and will cause a warning. The replacement for the ``nodelay`` short-form boolean option is ``nodelay=on`` rather than ``delay=off``. -``-smp`` ("parameter=0" SMP configurations) (since 6.2) -''''''''''''''''''''''''''''''''''''''''''''''''''''''' - -Specified CPU topology parameters must be greater than zero. - -In the SMP configuration, users should either provide a CPU topology -parameter with a reasonable value (greater than zero) or just omit it -and QEMU will compute the missing value. - -However, historically it was implicitly allowed for users to provide -a parameter with zero value, which is meaningless and could also possibly -cause unexpected results in the -smp parsing. So support for this kind of -configurations (e.g. -smp 8,sockets=0) is deprecated since 6.2 and will -be removed in the near future, users have to ensure that all the topology -members described with -smp are greater than zero. - Plugin argument passing through ``arg=<string>`` (since 6.1) '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''' diff --git a/docs/about/removed-features.rst b/docs/about/removed-features.rst index 417a0e4fa1..f9cf874f7b 100644 --- a/docs/about/removed-features.rst +++ b/docs/about/removed-features.rst @@ -489,6 +489,21 @@ The ``-singlestep`` option has been turned into an accelerator property, and given a name that better reflects what it actually does. Use ``-accel tcg,one-insn-per-tb=on`` instead. +``-smp`` ("parameter=0" SMP configurations) (removed in 9.0) +'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''' + +Specified CPU topology parameters must be greater than zero. + +In the SMP configuration, users should either provide a CPU topology +parameter with a reasonable value (greater than zero) or just omit it +and QEMU will compute the missing value. + +However, historically it was implicitly allowed for users to provide +a parameter with zero value, which is meaningless and could also possibly +cause unexpected results in the -smp parsing. So support for this kind of +configurations (e.g. -smp 8,sockets=0) is removed since 9.0, users have +to ensure that all the topology members described with -smp are greater +than zero. User-mode emulator command line arguments ----------------------------------------- diff --git a/hw/core/machine-smp.c b/hw/core/machine-smp.c index 25019c91ee..96533886b1 100644 --- a/hw/core/machine-smp.c +++ b/hw/core/machine-smp.c @@ -105,8 +105,9 @@ void machine_parse_smp_config(MachineState *ms, (config->has_cores && config->cores == 0) || (config->has_threads && config->threads == 0) || (config->has_maxcpus && config->maxcpus == 0)) { - warn_report("Deprecated CPU topology (considered invalid): " - "CPU topology parameters must be greater than zero"); + error_setg(errp, "Invalid CPU topology: " + "CPU topology parameters must be greater than zero"); + return; } /* -- 2.41.0

1 0

RE: join running core dump job
by Thanos Makatos 08 Mar '24

08 Mar '24

> -----Original Message----- > From: Thanos Makatos <thanos.makatos(a)nutanix.com> > Sent: Monday, March 4, 2024 9:45 PM > To: Thanos Makatos <thanos.makatos(a)nutanix.com>; devel(a)lists.libvirt.org > Subject: RE: join running core dump job > > > -----Original Message----- > > From: Thanos Makatos <thanos.makatos(a)nutanix.com> > > Sent: Monday, March 4, 2024 5:24 PM > > To: devel(a)lists.libvirt.org > > Subject: join running core dump job > > > > Is there a way to programmatically wait for a previously initiated > > virDomainCoreDumpWithFormat() where the process that started it died? > I'm > > looking at the API and don't seem to find anything relevant. I suppose I > could > > poll via virDomainGetJobStats(), but, ideally, I'd like a function that would > join > > the dump job and return when the dump job finishes. > > _______________________________________________ > > Devel mailing list -- devel(a)lists.libvirt.org > > To unsubscribe send an email to devel-leave(a)lists.libvirt.org > > I see there's qemuDumpWaitForCompletion(), looks promising. I've made some progress (added a virHypervisorDriver.domainCoreDumpWait and the relevant scaffolding to make 'virsh dump --wait' work), calling qemuDumpWaitForCompletion() is all that's needed. However, it doesn't seem trivial to implement this in the test_driver. First, IIUC testDomainCoreDumpWithFormat() gets an exclusive lock on the domain (haven't tested anything yet), so calling domainCoreDumpWait() would block for the wrong reason. Is making testDomainCoreDumpWithFormat() using asynchronous jobs internally unavoidable? Second, I want to test behaviour in my application where (1) it calls domainCoreDump(), (2) crashes before domainCoreDump() finishes, (3) then my application starts again and looks for a pending dump job and (4) joins it using domainCoreDumpWait(). I can't see an easy wait of faking a dump job in the test_driver when it starts. How about adding persistent tasks, which I can pre-populate before starting my application, or fake jobs via an environment variable, so that when the test_driver starts it can internally continue them? E.g. we can specify how long to run the job for and domainCoreDumpWait() add a sleep for that long. I'm open to suggestions.

1 0

[PATCH 0/2] ch_driver: add basic save and restore functionalities
by Purna Pavan Chandra Aekkaladevi 08 Mar '24

08 Mar '24

Purna Pavan Chandra Aekkaladevi (2): ch_driver: Add basic domain save & restore support NEWS: Mention save & restore support for ch driver NEWS.rst | 6 + src/ch/ch_conf.c | 6 + src/ch/ch_conf.h | 12 ++ src/ch/ch_driver.c | 511 +++++++++++++++++++++++++++++++++++++++++++- src/ch/ch_monitor.c | 97 ++++++++- src/ch/ch_monitor.h | 6 +- src/ch/ch_process.c | 101 +++++++-- src/ch/ch_process.h | 4 + 8 files changed, 721 insertions(+), 22 deletions(-) -- 2.34.1

2 4

[libvirt PATCH 0/3] acouple of net-metadata fixes
by Ján Tomko 08 Mar '24

08 Mar '24

Ján Tomko (3): remote: add VIR_ERR_NO_NETWORK_METADATA to daemonErrorLogFilter virsh: remove trailing whitespace even when editing the description vsh: introduce vshEditString src/remote/remote_daemon.c | 1 + tools/virsh-domain.c | 21 +-------------------- tools/virsh-network.c | 20 +------------------- tools/vsh.c | 27 +++++++++++++++++++++++++++ tools/vsh.h | 1 + 5 files changed, 31 insertions(+), 39 deletions(-) -- 2.43.2

2 6

[libvirt PATCH V3 0/4] add loongarch support for libvirt
by Xianglai Li 07 Mar '24

07 Mar '24

Hello, Everyone: This patch series adds libvirt support for loongarch.Although the bios path and name has not been officially integrated into qemu and we think there are still many shortcomings, we try to push a version of patch to the community according to the opinions of the community, hoping to listen to everyone's opinions. Anyway we have a version of libvirt that supports loongarch. You can also get libvirt's patch from the link below: https://gitlab.com/lixianglai/libvirt branch: loongarch Since the patch associated with loongarch has not yet been submitted to the virt-manager community, we are providing a temporary patch with loongarch for the time being patch's virt-manager, the open source work of virt-manager adding loongarch will be followed up later or synchronized with the open source libvirt. You can get the virt-manager code with loongarch patch from the link below: https://github.com/lixianglai/virt-manager branch: loongarch loongarch's virtual machine bios is not yet available in qemu,So you need to compile loongarch UEFI yourself, you can refer to the following link to compile UEFI: https://github.com/tianocore/edk2-platforms/blob/master/Platform/Loongson/L… Here we provide compiled UEFI for ease of testing,you can get it from the following link: https://github.com/lixianglai/LoongarchVirtFirmware (Note: You should clone the repository using git instead of downloading the file via wget or you'll get xml) We named the bios QEMU_EFI.fd, QEMU_VARS.fd is used to store pflash images of non-volatile variables. After installing qemu-system-loongarch64,You can install the loongarch bios by executing the script install-loongarch-virt-firmware.sh Since there is no fedora operating system that supports the loongarch architecture, you can find an iso that supports loongarch at the link below for testing purposes: https://github.com/fedora-remix-loongarch/releases-info The operating system provided above may have minor problems of one kind or another, but you can also find out from the following link that openEuler's operating system supports loongarch: https://muug.ca/mirror/openeuler/openEuler-22.03-LTS/ISO/loongarch64/ Well, if you have completed the above steps I think you can now install loongarch virtual machine, you can install it through the virt-manager graphical interface, or install it through vrit-install, here is an example of installing it using virt-install: virt-install \ --virt-type=qemu \ --name loongarch-test \ --memory 4096 \ --vcpus=4 \ --arch=loongarch64 \ --boot cdrom \ --disk device=cdrom,bus=scsi,path=/root/livecd-fedora-mate-4.loongarch64.iso \ --disk path=/var/lib/libvirt/images/debian12-loongarch64.qcow2,size=10,format=qcow2,bus=scsi \ --network network=default \ --osinfo archlinux \ --video=virtio \ --graphics=vnc,listen=0.0.0.0 CHANGES V2->V3: 1.Delete redundant header files in cpu_loongarch.c 2.Fixed code formatting issues 3.Modify the commit message of [PATCH 2/4] 4.Remove useless test cases for loongarch, Rebuild loongarch's test case based on the latest code. V1->V2: 1.Modify the link addresses of virtu-manager and firmeware in the cover letter. Please obtain the code and firmware from the latest link address. 2.Rename the bios name. Delete the loongarch bios name from libvirt and use the json file to obtain the bios path. 3.Refer to riscv64 to simplify the implementation of loongarch cpu driver.And fix some code style errors. 4.Delete unnecessary or redundant device enablement.Such as USB NEC. 5.Add some test cases for loongarch. Xianglai Li (4): Add loongarch cpu support Support for loongarch64 in the QEMU driver Implement the method of getting host info for loongarch Add test script for loongarch src/conf/schemas/basictypes.rng | 1 + src/cpu/cpu.c | 2 + src/cpu/cpu_loongarch.c | 58 + src/cpu/cpu_loongarch.h | 25 + src/cpu/meson.build | 1 + src/qemu/qemu_capabilities.c | 16 +- src/qemu/qemu_command.c | 7 +- src/qemu/qemu_domain.c | 40 +- src/qemu/qemu_domain.h | 1 + src/qemu/qemu_validate.c | 1 + src/util/virarch.c | 13 +- src/util/virarch.h | 13 +- src/util/virhostcpu.c | 7 +- src/util/virsysinfo.c | 3 +- .../qemu_8.2.0-tcg-virt.loongarch64.xml | 163 + .../qemu_8.2.0-virt.loongarch64.xml | 167 + tests/domaincapstest.c | 4 +- .../caps_8.2.0_loongarch64.replies | 30121 ++++++++++++++++ .../caps_8.2.0_loongarch64.xml | 175 + .../qemucaps2xmloutdata/caps.loongarch64.xml | 28 + ...o-type-loongarch64.loongarch64-latest.args | 34 + ...eo-type-loongarch64.loongarch64-latest.xml | 45 + .../default-video-type-loongarch64.xml | 18 + ...-models.loongarch64-latest.abi-update.args | 44 + ...t-models.loongarch64-latest.abi-update.xml | 79 + ...irt-default-models.loongarch64-latest.args | 44 + ...virt-default-models.loongarch64-latest.xml | 79 + .../loongarch64-virt-default-models.xml | 24 + ...ch64-virt-graphics.loongarch64-latest.args | 56 + ...rch64-virt-graphics.loongarch64-latest.xml | 116 + .../loongarch64-virt-graphics.xml | 48 + ...ch64-virt-headless.loongarch64-latest.args | 52 + ...rch64-virt-headless.loongarch64-latest.xml | 102 + .../loongarch64-virt-headless.xml | 42 + ...minimal.loongarch64-latest.abi-update.args | 31 + ...-minimal.loongarch64-latest.abi-update.xml | 26 + ...rch64-virt-minimal.loongarch64-latest.args | 31 + ...arch64-virt-minimal.loongarch64-latest.xml | 26 + .../loongarch64-virt-minimal.xml | 15 + tests/qemuxmlconftest.c | 7 + tests/testutilshostcpus.h | 10 + 41 files changed, 31749 insertions(+), 26 deletions(-) create mode 100644 src/cpu/cpu_loongarch.c create mode 100644 src/cpu/cpu_loongarch.h create mode 100644 tests/domaincapsdata/qemu_8.2.0-tcg-virt.loongarch64.xml create mode 100644 tests/domaincapsdata/qemu_8.2.0-virt.loongarch64.xml create mode 100644 tests/qemucapabilitiesdata/caps_8.2.0_loongarch64.replies create mode 100644 tests/qemucapabilitiesdata/caps_8.2.0_loongarch64.xml create mode 100644 tests/qemucaps2xmloutdata/caps.loongarch64.xml create mode 100644 tests/qemuxmlconfdata/default-video-type-loongarch64.loongarch64-latest.args create mode 100644 tests/qemuxmlconfdata/default-video-type-loongarch64.loongarch64-latest.xml create mode 100644 tests/qemuxmlconfdata/default-video-type-loongarch64.xml create mode 100644 tests/qemuxmlconfdata/loongarch64-virt-default-models.loongarch64-latest.abi-update.args create mode 100644 tests/qemuxmlconfdata/loongarch64-virt-default-models.loongarch64-latest.abi-update.xml create mode 100644 tests/qemuxmlconfdata/loongarch64-virt-default-models.loongarch64-latest.args create mode 100644 tests/qemuxmlconfdata/loongarch64-virt-default-models.loongarch64-latest.xml create mode 100644 tests/qemuxmlconfdata/loongarch64-virt-default-models.xml create mode 100644 tests/qemuxmlconfdata/loongarch64-virt-graphics.loongarch64-latest.args create mode 100644 tests/qemuxmlconfdata/loongarch64-virt-graphics.loongarch64-latest.xml create mode 100644 tests/qemuxmlconfdata/loongarch64-virt-graphics.xml create mode 100644 tests/qemuxmlconfdata/loongarch64-virt-headless.loongarch64-latest.args create mode 100644 tests/qemuxmlconfdata/loongarch64-virt-headless.loongarch64-latest.xml create mode 100644 tests/qemuxmlconfdata/loongarch64-virt-headless.xml create mode 100644 tests/qemuxmlconfdata/loongarch64-virt-minimal.loongarch64-latest.abi-update.args create mode 100644 tests/qemuxmlconfdata/loongarch64-virt-minimal.loongarch64-latest.abi-update.xml create mode 100644 tests/qemuxmlconfdata/loongarch64-virt-minimal.loongarch64-latest.args create mode 100644 tests/qemuxmlconfdata/loongarch64-virt-minimal.loongarch64-latest.xml create mode 100644 tests/qemuxmlconfdata/loongarch64-virt-minimal.xml -- 2.39.1

4 11

[PATCH 0/2] Add support for Unix mode to serial port dev
by Praveen K Paladugu 07 Mar '24

07 Mar '24

This patchset adds support for Unix mode to serial port device in ch guests. Praveen K Paladugu (2): ch: Add Cap checks for unix backend of serial port ch: Add support for `Unix` mode to serial port dev src/ch/ch_capabilities.c | 6 ++++++ src/ch/ch_capabilities.h | 1 + src/ch/ch_domain.c | 21 +++++++++++++++------ src/ch/ch_monitor.c | 35 ++++++++++++++++++++++++----------- src/ch/ch_process.c | 7 +++++++ 5 files changed, 53 insertions(+), 17 deletions(-) -- 2.43.0

2 3

[PATCH] tests: Don't advertise features for edk2-loongarch64
by Andrea Bolognani 07 Mar '24

07 Mar '24

The "acpi" feature is not part of the specification[1], so it will just get ignored. [1] https://gitlab.com/qemu-project/qemu/-/blob/master/docs/interop/firmware.js… Signed-off-by: Andrea Bolognani <abologna(a)redhat.com> --- .../usr/share/qemu/firmware/50-edk2-loongarch64.json | 1 - 1 file changed, 1 deletion(-) diff --git a/tests/qemufirmwaredata/usr/share/qemu/firmware/50-edk2-loongarch64.json b/tests/qemufirmwaredata/usr/share/qemu/firmware/50-edk2-loongarch64.json index 8251137ba9..c5a7ec5f7b 100644 --- a/tests/qemufirmwaredata/usr/share/qemu/firmware/50-edk2-loongarch64.json +++ b/tests/qemufirmwaredata/usr/share/qemu/firmware/50-edk2-loongarch64.json @@ -24,6 +24,5 @@ } ], "features": [ - "acpi" ] } -- 2.44.0

2 1

Raspberry Pi 5 suggestion
by K otgc 07 Mar '24

07 Mar '24

Hello, the libvirt documentation <https://libvirt.org/apps.html> has nothing on installing libvirt to Raspberry Pi 5, so the documentation says to email <https://libvirt.org/contact.html> any suggestions. Cheers, Ken.

3 3

[PATCH] docs: Update documentation of CPU models in domain caps
by Jiri Denemark 07 Mar '24

07 Mar '24

Using check='none' when starting a domain with a CPU model marked as usable is no longer needed as libvirt will do the right thing even with check='partial'. Signed-off-by: Jiri Denemark <jdenemar(a)redhat.com> --- docs/formatdomaincaps.rst | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/docs/formatdomaincaps.rst b/docs/formatdomaincaps.rst index ef752a0f3a..1999112944 100644 --- a/docs/formatdomaincaps.rst +++ b/docs/formatdomaincaps.rst @@ -237,10 +237,12 @@ more details about it: using the CPU model and no additional feature elements. Models marked as usable (``usable='yes'``) can be safely used in domain XMLs with ``check='none'`` as the hypervisor guarantees the model can be used on the - current host and additional checks done by libvirt are redundant. In fact, - disabling libvirt checks via ``check='none'`` for such models is recommended - to avoid needless issues with starting domains when libvirt's definition of - a particular model differs from hypervisor's definition. The + current host and additional checks done by libvirt are redundant. + :since:`Since 10.2.0` libvirt automatically detects this situation and + avoids the redundant checks, with older releases disabling libvirt checks + via ``check='none'`` for such models is recommended to avoid needless + issues with starting domains when libvirt's definition of a particular model + differs from hypervisor's definition. The ``deprecated`` attribute reflects the hypervisor's policy on usage of this model :since:`(since 7.1.0)`. The ``vendor`` attribute :since:`(since 8.9.0)` contains the vendor of the CPU model for users who want to use CPU models -- 2.44.0

2 1

[PATCH 02/14] hw/core/machine-smp: Deprecate unsupported "parameter=1" SMP configurations
by Zhao Liu 07 Mar '24

07 Mar '24

From: Zhao Liu <zhao1.liu(a)intel.com> Currentlt, it was allowed for users to specify the unsupported topology parameter as "1". For example, x86 PC machine doesn't support drawer/book/cluster topology levels, but user could specify "-smp drawers=1,books=1,clusters=1". This is meaningless and confusing, so that the support for this kind of configurations is marked depresated since 9.0. And report warning message for such case like: qemu-system-x86_64: warning: Deprecated CPU topology (considered invalid): Unsupported clusters parameter mustn't be specified as 1 qemu-system-x86_64: warning: Deprecated CPU topology (considered invalid): Unsupported books parameter mustn't be specified as 1 qemu-system-x86_64: warning: Deprecated CPU topology (considered invalid): Unsupported drawers parameter mustn't be specified as 1 Users have to ensure that all the topology members described with -smp are supported by the target machine. Cc: devel(a)lists.libvirt.org Signed-off-by: Zhao Liu <zhao1.liu(a)intel.com> --- docs/about/deprecated.rst | 14 +++++++++ hw/core/machine-smp.c | 63 +++++++++++++++++++++++++++++---------- 2 files changed, 61 insertions(+), 16 deletions(-) diff --git a/docs/about/deprecated.rst b/docs/about/deprecated.rst index 872974640252..2e782e83e952 100644 --- a/docs/about/deprecated.rst +++ b/docs/about/deprecated.rst @@ -47,6 +47,20 @@ as short-form boolean values, and passed to plugins as ``arg_name=on``. However, short-form booleans are deprecated and full explicit ``arg_name=on`` form is preferred. +``-smp`` (Unsopported "parameter=1" SMP configurations) (since 9.0) +''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''' + +Specified CPU topology parameters must be supported by the machine. + +In the SMP configuration, users should provide the CPU topology parameters that +are supported by the target machine. + +However, historically it was allowed for users to specify the unsupported +topology parameter as "1", which is meaningless. So support for this kind of +configurations (e.g. -smp drawers=1,books=1,clusters=1 for x86 PC machine) is +marked depresated since 9.0, users have to ensure that all the topology members +described with -smp are supported by the target machine. + QEMU Machine Protocol (QMP) commands ------------------------------------ diff --git a/hw/core/machine-smp.c b/hw/core/machine-smp.c index 96533886b14e..50a5a40dbc3d 100644 --- a/hw/core/machine-smp.c +++ b/hw/core/machine-smp.c @@ -112,30 +112,61 @@ void machine_parse_smp_config(MachineState *ms, /* * If not supported by the machine, a topology parameter must be - * omitted or specified equal to 1. + * omitted. */ - if (!mc->smp_props.dies_supported && dies > 1) { - error_setg(errp, "dies not supported by this machine's CPU topology"); - return; - } - if (!mc->smp_props.clusters_supported && clusters > 1) { - error_setg(errp, "clusters not supported by this machine's CPU topology"); - return; + if (!mc->smp_props.clusters_supported && config->has_clusters) { + if (config->clusters > 1) { + error_setg(errp, "clusters not supported by this " + "machine's CPU topology"); + return; + } else { + /* Here clusters only equals 1 since we've checked zero case. */ + warn_report("Deprecated CPU topology (considered invalid): " + "Unsupported clusters parameter mustn't be " + "specified as 1"); + } } + clusters = clusters > 0 ? clusters : 1; + if (!mc->smp_props.dies_supported && config->has_dies) { + if (config->dies > 1) { + error_setg(errp, "dies not supported by this " + "machine's CPU topology"); + return; + } else { + /* Here dies only equals 1 since we've checked zero case. */ + warn_report("Deprecated CPU topology (considered invalid): " + "Unsupported dies parameter mustn't be " + "specified as 1"); + } + } dies = dies > 0 ? dies : 1; - clusters = clusters > 0 ? clusters : 1; - if (!mc->smp_props.books_supported && books > 1) { - error_setg(errp, "books not supported by this machine's CPU topology"); - return; + if (!mc->smp_props.books_supported && config->has_books) { + if (config->books > 1) { + error_setg(errp, "books not supported by this " + "machine's CPU topology"); + return; + } else { + /* Here books only equals 1 since we've checked zero case. */ + warn_report("Deprecated CPU topology (considered invalid): " + "Unsupported books parameter mustn't be " + "specified as 1"); + } } books = books > 0 ? books : 1; - if (!mc->smp_props.drawers_supported && drawers > 1) { - error_setg(errp, - "drawers not supported by this machine's CPU topology"); - return; + if (!mc->smp_props.drawers_supported && config->has_drawers) { + if (config->drawers > 1) { + error_setg(errp, "drawers not supported by this " + "machine's CPU topology"); + return; + } else { + /* Here drawers only equals 1 since we've checked zero case. */ + warn_report("Deprecated CPU topology (considered invalid): " + "Unsupported drawers parameter mustn't be " + "specified as 1"); + } } drawers = drawers > 0 ? drawers : 1; -- 2.34.1

2 2