Few fixes I found when building QEMU with Rust
cover Few fixes I found when building QEMU with Rust
I'm fairly sure the first one is alright, but I must admit I have no idea why is the second patch needed on my system.
For both patches, please take them with a grain of salt and feel free to prove me wrong with anything.
This used to be a bigger series until I rebased on top of Paolo's rust-next branch and basically got half of it for free. That's why these unrelated patches are bundled in a "series".
Martin Kletzander (2):
rust: Add antoher variant for impl_vmstate_struct! macro tests/qemu-iotests: Indent expected error messages
rust/qemu-api/src/vmstate.rs | 11 +++++++++++ rust/qemu-api/tests/vmstate_tests.rs | 1 + tests/qemu-iotests/039.out | 10 +++++----- tests/qemu-iotests/061.out | 4 ++-- tests/qemu-iotests/137.out | 2 +- 5 files changed, 20 insertions(+), 8 deletions(-)
Signature · 1 line
2.50.1
PATCH 2/2 tests/qemu-iotests: Indent expected error messages
From: Martin Kletzander <[email protected]>
When running all tests the expected "killed" messages are indented differently than the actual ones, by three more spaces. Change it so that the messages match and tests pass.
--- tests/qemu-iotests/039.out | 10 +++++----- tests/qemu-iotests/061.out | 4 ++-- tests/qemu-iotests/137.out | 2 +- 3 files changed, 8 insertions(+), 8 deletions(-)
tests/qemu-iotests/039.out +5 -5
@@ -11,7 +11,7 @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) incompatible_features [0] ERROR cluster 5 refcount=0 reference=1 ERROR OFLAG_COPIED data cluster: l2_entry=8000000000050000 refcount=0
@@ -46,7 +46,7 @@ read 512/512 bytes at offset 0 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) incompatible_features [0] ERROR cluster 5 refcount=0 reference=1 Rebuilding refcount structure
@@ -60,7 +60,7 @@ incompatible_features [] Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) incompatible_features [] No errors were found on the image.
@@ -79,7 +79,7 @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) incompatible_features [0] ERROR cluster 5 refcount=0 reference=1 ERROR OFLAG_COPIED data cluster: l2_entry=8000000000050000 refcount=0
@@ -89,7 +89,7 @@ Data may be corrupted, or further writes to the image may corrupt it. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) incompatible_features [] No errors were found on the image. *** done
tests/qemu-iotests/061.out +2 -2
@@ -118,7 +118,7 @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 wrote 131072/131072 bytes at offset 0 128 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) magic 0x514649fb version 3 backing_file_offset 0x0
@@ -304,7 +304,7 @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 wrote 131072/131072 bytes at offset 0 128 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) magic 0x514649fb version 3 backing_file_offset 0x0
tests/qemu-iotests/137.out +1 -1
@@ -35,7 +35,7 @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 qemu-io: Unsupported value 'blubb' for qcow2 option 'overlap-check'. Allowed are any of the following: none, constant, cached, all wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) OK: Dirty bit not set Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 qemu-io: Parameter 'lazy-refcounts' expects 'on' or 'off'
Signature · 1 line
2.50.1
PATCH 1/2 rust: Add antoher variant for impl_vmstate_struct! macro
In reply to Martin KletzanderFrom: Martin Kletzander <[email protected]>
In some cases (e.g. in vmstate_tests.rs) the second argument to impl_vmstate_struct! is actually an existing struct which is then copied (since VMStateDescription implements Copy) when saved into the static VMSD using .get(). That is not a problem because it is part of the data segment and the pointers are not being free'd since they point to static data. But it is a problem when tests rely on comparing the VMState descriptions as pointers rather than contents. And it also wastes space, more or less.
Introduce second variant of the macro which can, instead of the expression, take an identifier or what looks like a reference. This second variant is added before the current variant so that it has preference, and only references the existing static data from it.
This way tests are fixed and space is saved.
And now that the VMStateDescription checking is fixed we can also check for the right value in test_vmstate_struct_varray_uint8_wrapper().
---
I'm not sure whether this is caused by different utility on my system or bash version or whatever, but without this patch these three tests fail for me and this patch fixes it.
rust/qemu-api/src/vmstate.rs | 11 +++++++++++ rust/qemu-api/tests/vmstate_tests.rs | 1 + 2 files changed, 12 insertions(+)
rust/qemu-api/src/vmstate.rs +11 -0
@@ -449,6 +449,17 @@ macro_rules! vmstate_validate {/// description of the struct. #[macro_export] macro_rules! impl_vmstate_struct { + ($type:ty, $(&)?$vmsd:ident) => { +unsafe impl $crate::vmstate::VMState for $type { +const BASE : $crate::bindings::VMStateField = + $crate::bindings::VMStateField { + vmsd: $vmsd.as_ref (), + size: ::core::mem::size_of ::<$type>(), + flags: $crate::bindings::VMStateFlags ::VMS_STRUCT , + ..$crate::zeroable::Zeroable ::ZERO + }; + } + }; ($type:ty, $vmsd:expr) => {unsafe impl $crate::vmstate::VMState for $type {const BASE : $crate::bindings::VMStateField = {
rust/qemu-api/tests/vmstate_tests.rs +1 -0
@@ -320,6 +320,7 @@ fn test_vmstate_struct_varray_uint8_wrapper() {b"arr_a_wrap\0" );assert_eq! (foo_fields[5 ].num_offset,228 ); +assert_eq! (foo_fields[5 ].vmsd,VMSTATE_FOOA .as_ref ());assert! (unsafe { foo_fields[5 ].field_exists.unwrap ()(foo_b_p,0 ) });// The last VMStateField in VMSTATE_FOOB.
Signature · 1 line
2.50.1
Re: [PATCH 2/2] tests/qemu-iotests: Indent expected error messages
In reply to Martin KletzanderQuoted 5 lines · Martin Kletzander
From: Martin Kletzander <[email protected]>
When running all tests the expected "killed" messages are indented differently than the actual ones, by three more spaces. Change it so that the messages match and tests pass.
This would break the tests on my system and CI too.
What distro are you seeing this on ?
I'm guessing this is a different in either valgrind or C library ?
Quoted 96 lines
Signed-off-by Martin Kletzander <[email protected]>--- tests/qemu-iotests/039.out | 10 +++++----- tests/qemu-iotests/061.out | 4 ++-- tests/qemu-iotests/137.out | 2 +- 3 files changed, 8 insertions(+), 8 deletions(-)tests/qemu-iotests/039.out +5
-5@@ -11,7 +11,7 @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) incompatible_features [0] ERROR cluster 5 refcount=0 reference=1 ERROR OFLAG_COPIED data cluster: l2_entry=8000000000050000 refcount=0@@ -46,7 +46,7 @@ read 512/512 bytes at offset 0 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) incompatible_features [0] ERROR cluster 5 refcount=0 reference=1 Rebuilding refcount structure@@ -60,7 +60,7 @@ incompatible_features [] Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) incompatible_features [] No errors were found on the image.@@ -79,7 +79,7 @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) incompatible_features [0] ERROR cluster 5 refcount=0 reference=1 ERROR OFLAG_COPIED data cluster: l2_entry=8000000000050000 refcount=0@@ -89,7 +89,7 @@ Data may be corrupted, or further writes to the image may corrupt it. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) incompatible_features [] No errors were found on the image. *** donetests/qemu-iotests/061.out +2
-2@@ -118,7 +118,7 @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 wrote 131072/131072 bytes at offset 0 128 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) magic 0x514649fb version 3 backing_file_offset 0x0@@ -304,7 +304,7 @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 wrote 131072/131072 bytes at offset 0 128 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) magic 0x514649fb version 3 backing_file_offset 0x0tests/qemu-iotests/137.out +1
-1@@ -35,7 +35,7 @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 qemu-io: Unsupported value 'blubb' for qcow2 option 'overlap-check'. Allowed are any of the following: none, constant, cached, all wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) OK: Dirty bit not set Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 qemu-io: Parameter 'lazy-refcounts' expects 'on' or 'off'Signature · 1 line
2.50.1
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 2/2] tests/qemu-iotests: Indent expected error messages
Quoted 7 lines · Daniel P. Berrangé
Quoted 5 lines · Martin Kletzander
From: Martin Kletzander <[email protected]>
When running all tests the expected "killed" messages are indented differently than the actual ones, by three more spaces. Change it so that the messages match and tests pass.
Quoted 6 lines
This would break the tests on my system and CI too.
What distro are you seeing this on ?
I'm guessing this is a different in either valgrind or C library ?
It's bash, we have an open issue about it:
Re: [PATCH 1/2] rust: Add antoher variant for impl_vmstate_struct! macro
In reply to Martin KletzanderIl ven 1 ago 2025, 17:00 Martin Kletzander <[email protected]> ha scritto:
Quoted 21 lines
From: Martin Kletzander <[email protected]>
In some cases (e.g. in
vmstate_tests.rs) the second argument toimpl_vmstate_struct! is actually an existing struct which is then copied (since VMStateDescription implements Copy) when saved into the static VMSD using .get(). That is not a problem because it is part of the data segment and the pointers are not being free'd since they point to static data. But it is a problem when tests rely on comparing the VMState descriptions as pointers rather than contents. And it also wastes space, more or less.Introduce second variant of the macro which can, instead of the expression, take an identifier or what looks like a reference. This second variant is added before the current variant so that it has preference, and only references the existing static data from it.
This way tests are fixed and space is saved.
And now that the VMStateDescription checking is fixed we can also check for the right value in
test_vmstate_struct_varray_uint8_wrapper().
Quoted 8 lines
Signed-off-by Martin Kletzander <[email protected]>---
I'm not sure whether this is caused by different utility on my system or bash
version or whatever, but without this patch these three tests fail for me and
this patch fixes it.
I found something similar, though I wasn't sure if it was broken in master as well or only in the rust-next branch.
If that works in master as well, I would remove completely the possibility of using &FOO, and always use .as_ref(). It's more efficient as you said, and there's no reason that I know to use the less efficient one.
Paolo
rust/qemu-api/src/vmstate.rs | 11 +++++++++++
Quoted 42 lines
rust/qemu-api/tests/vmstate_tests.rs | 1 + 2 files changed, 12 insertions(+)rust/qemu-api/src/vmstate.rs +11
-0@@ -449,6 +449,17 @@ macro_rules! vmstate_validate {/// description of the struct. #[macro_export] macro_rules! impl_vmstate_struct { + ($type:ty, $(&)?$vmsd:ident) => { +unsafe impl $crate::vmstate::VMState for $type { +const BASE : $crate::bindings::VMStateField = + $crate::bindings::VMStateField { + vmsd: $vmsd.as_ref (), + size: ::core::mem::size_of ::<$type>(), + flags: $crate::bindings::VMStateFlags ::VMS_STRUCT , + ..$crate::zeroable::Zeroable ::ZERO + }; + } + }; ($type:ty, $vmsd:expr) => {unsafe impl $crate::vmstate::VMState for $type {const BASE : $crate::bindings::VMStateField = {rust/qemu-api/tests/ renamed from rust/qemu-api/tests/vmstate_tests.rs +0
-0
vmstate_tests.rs
index2c0670ba0eed..7d3180e6c2ea100644rust/qemu-api/tests/vmstate_tests.rs +1
-1@@ -320,6 +320,7 @@ fn test_vmstate_struct_varray_uint8_wrapper() {b"arr_a_wrap\0" );assert_eq! (foo_fields[5 ].num_offset,228 ); +assert_eq! (foo_fields[5 ].vmsd,VMSTATE_FOOA .as_ref ());assert! (unsafe { foo_fields[5 ].field_exists.unwrap ()(foo_b_p,0 ) });// The last VMStateField in VMSTATE_FOOB. --2.50.1
Re: [PATCH 2/2] tests/qemu-iotests: Indent expected error messages
In reply to Fabiano RosasQuoted 11 lines · Fabiano Rosas
Quoted 9 lines · Daniel P. Berrangé
Quoted 5 lines · Martin Kletzander
From: Martin Kletzander <[email protected]>
When running all tests the expected "killed" messages are indented differently than the actual ones, by three more spaces. Change it so that the messages match and tests pass.
This would break the tests on my system and CI too.
Quoted 9 lines
Quoted 4 lines
What distro are you seeing this on ?
I'm guessing this is a different in either valgrind or C library ?
It's bash, we have an open issue about it:
Ah, good to know, thanks. That's what I suspected, but haven't thought of checking gitlab issues. Sorry for the noise.
Re: [PATCH 1/2] rust: Add antoher variant for impl_vmstate_struct! macro
In reply to Paolo BonziniQuoted 32 lines · Paolo Bonzini
Il ven 1 ago 2025, 17:00 Martin Kletzander <[email protected]> ha scritto:
Quoted 21 lines
From: Martin Kletzander <[email protected]>
In some cases (e.g. in
vmstate_tests.rs) the second argument toimpl_vmstate_struct! is actually an existing struct which is then copied (since VMStateDescription implements Copy) when saved into the static VMSD using .get(). That is not a problem because it is part of the data segment and the pointers are not being free'd since they point to static data. But it is a problem when tests rely on comparing the VMState descriptions as pointers rather than contents. And it also wastes space, more or less.Introduce second variant of the macro which can, instead of the expression, take an identifier or what looks like a reference. This second variant is added before the current variant so that it has preference, and only references the existing static data from it.
This way tests are fixed and space is saved.
And now that the VMStateDescription checking is fixed we can also check for the right value in
test_vmstate_struct_varray_uint8_wrapper().Quoted 8 lines
Signed-off-by Martin Kletzander <[email protected]>---
I'm not sure whether this is caused by different utility on my system or bash
version or whatever, but without this patch these three tests fail for me and
this patch fixes it.
Quoted 3 lines
I found something similar, though I wasn't sure if it was broken in master as well or only in the rust-next branch.
It is not broken on master, but it *was* broken on rust-next.
Quoted 4 lines
If that works in master as well, I would remove completely the possibility of using &FOO, and always use .as_ref(). It's more efficient as you said, and there's no reason that I know to use the less efficient one.
That would mean that you cannot do what's done in e.g. pl011
impl_vmstate_struct!(
PL011Registers,
VMStateDescriptionBuilder::<PL011Registers>::new()
...
.build());
requiring you to always do:
static/const VMSTATE_PL011_REGISTERS: VMStateDescription<PL011Registers> =
VMStateDescriptionBuilder::<PL011Registers>::new()
...
.build();impl_vmstate_struct!(PL011Registers, VMSTATE_PL011_REGISTERS);
*BUT* of course I had to rebase the patches on top of current rust-next on Friday and there were some of your commits from Thursday which I now see actually fix all what I tried fixing before as well. I tried finding the previous commit on which I saw all the issues and after some rebuilding I could not. So it is now not even broken on rust-next.
This way I completely wasted your time, but at least learned something that's happening in the code. Sorry for that.
Martin
Quoted 45 lines
Paolo
rust/qemu-api/src/vmstate.rs| 11 +++++++++++Quoted 42 lines
rust/qemu-api/tests/vmstate_tests.rs | 1 + 2 files changed, 12 insertions(+)rust/qemu-api/src/vmstate.rs +11
-0@@ -449,6 +449,17 @@ macro_rules! vmstate_validate {/// description of the struct. #[macro_export] macro_rules! impl_vmstate_struct { + ($type:ty, $(&)?$vmsd:ident) => { +unsafe impl $crate::vmstate::VMState for $type { +const BASE : $crate::bindings::VMStateField = + $crate::bindings::VMStateField { + vmsd: $vmsd.as_ref (), + size: ::core::mem::size_of ::<$type>(), + flags: $crate::bindings::VMStateFlags ::VMS_STRUCT , + ..$crate::zeroable::Zeroable ::ZERO + }; + } + }; ($type:ty, $vmsd:expr) => {unsafe impl $crate::vmstate::VMState for $type {const BASE : $crate::bindings::VMStateField = {rust/qemu-api/tests/ renamed from rust/qemu-api/tests/vmstate_tests.rs +0
-0
vmstate_tests.rs
index2c0670ba0eed..7d3180e6c2ea100644rust/qemu-api/tests/vmstate_tests.rs +1
-1@@ -320,6 +320,7 @@ fn test_vmstate_struct_varray_uint8_wrapper() {b"arr_a_wrap\0" );assert_eq! (foo_fields[5 ].num_offset,228 ); +assert_eq! (foo_fields[5 ].vmsd,VMSTATE_FOOA .as_ref ());assert! (unsafe { foo_fields[5 ].field_exists.unwrap ()(foo_b_p,0 ) });// The last VMStateField in VMSTATE_FOOB. --2.50.1
Re: [PATCH 1/2] rust: Add antoher variant for impl_vmstate_struct! macro
Il lun 4 ago 2025, 10:56 Martin Kletzander <[email protected]> ha scritto:
Quoted 9 lines
*BUT* of course I had to rebase the patches on top of current rust-next on Friday and there were some of your commits from Thursday which I now see actually fix all what I tried fixing before as well. I tried finding the previous commit on which I saw all the issues and after some rebuilding I could not. So it is now not even broken on rust-next.
This way I completely wasted your time, but at least learned something that's happening in the code. Sorry for that.
Uh no you didn't. It was broken.
Paolo
Martin
Quoted 50 lines
Quoted 8 lines
Paolo
rust/qemu-api/src/vmstate.rs| 11 +++++++++++rust/qemu-api/src/
Quoted 4 lines
rust/qemu-api/tests/vmstate_tests.rs | 1 + 2 files changed, 12 insertions(+)rust/qemu-api/src/ renamed from rust/qemu-api/src/vmstate.rs +0
-0
vmstate.rsQuoted 14 lines
Quoted 14 lines
index
b5c6b764fbba..716e52afe740100644rust/qemu-api/src/vmstate.rs +7
-0@@ -449,6 +449,17 @@ macro_rules! vmstate_validate {/// description of the struct. #[macro_export] macro_rules! impl_vmstate_struct { + ($type:ty, $(&)?$vmsd:ident) => { +unsafe impl $crate::vmstate::VMState for $type { +const BASE : $crate::bindings::VMStateField = + $crate::bindings::VMStateField { + vmsd: $vmsd.as_ref (), + size: ::core::mem::size_of ::<$type>(), + flags: $crate::bindings::VMStateFlags ::VMS_STRUCT ,Quoted 8 lines
Review
Quoted 8 lines
rust/qemu-api/src/vmstate.rs +4
-0+ ..$crate::zeroable::Zeroable ::ZERO + }; + } + }; ($type:ty, $vmsd:expr) => {unsafe impl $crate::vmstate::VMState for $type {const BASE : $crate::bindings::VMStateField = {(quoted diff) +0
-0b/rust/qemu-api/tests/
Quoted 10 lines
rust/qemu-api/tests/vmstate_tests.rs fn test_vmstate_struct_varray_uint8_wrapper() { · line 324
Quoted 10 lines
vmstate_tests.rs
index2c0670ba0eed..7d3180e6c2ea100644rust/qemu-api/tests/vmstate_tests.rs +1
-0@@ -320,6 +320,7 @@ fn test_vmstate_struct_varray_uint8_wrapper() {b"arr_a_wrap\0" );assert_eq! (foo_fields[5 ].num_offset,228 ); +assert_eq! (foo_fields[5 ].vmsd,VMSTATE_FOOA .as_ref ());assert! (unsafe { foo_fields[5 ].field_exists.unwrap ()(foo_b_p,0 )});
Quoted 6 lines
Quoted 6 lines
rust/qemu-api/tests/vmstate_tests.rs +0
-1// The last VMStateField in VMSTATE_FOOB. --2.50.1
Re: [PATCH 2/2] tests/qemu-iotests: Indent expected error messages
In reply to Fabiano RosasAm 01.08.2025 um 21:09 hat Fabiano Rosas geschrieben:
Quoted 11 lines
Quoted 9 lines · Daniel P. Berrangé
Quoted 5 lines · Martin Kletzander
From: Martin Kletzander <[email protected]>
When running all tests the expected "killed" messages are indented differently than the actual ones, by three more spaces. Change it so that the messages match and tests pass.
This would break the tests on my system and CI too.
Quoted 7 lines
Quoted 3 lines
What distro are you seeing this on ?
I'm guessing this is a different in either valgrind or C library ?
It's bash, we have an open issue about it:
I see a patch has been posted to that bug, the most important part of which is this added filtering:
tests/qemu-iotests/common.filter 2025-07-25 11:39:22.419665788 +0000 renamed from tests/qemu-iotests/common.filter +1 -1
@@ -75,7 +75,7 @@ _filter_qemu_io() _filter_win32 | \ gsed -e "s/[0-9]* ops\; [0-9/:. sec]* ([0-9/.inf]* [EPTGMKiBbytes]*\/sec and [0-9/.inf]* ops\/sec)/X ops\; XX:XX:XX.X (XXX YYY\/sec and XXX ops\/sec)/" \ -e "s/: line [0-9][0-9]*: *[0-9][0-9]*\( Aborted\| Killed\)/:\1/" \ - -e "s/qemu-io> //g" + -e "s/qemu-io> //g" -e '/Killed/{ s/ \{2,\}/ /}' } # replace occurrences of QEMU_PROG with "qemu"
This approach makes sense to me, though I would have kept each sed expression on a separate line.
And given that the context line above includes "Aborted" as well, maybe have it here, too, though none of the actual test outputs have an Aborted message any more since commit 3f39447. Or we could have a cleanup patch first that removes the unused "Abort" above, just to keep things consistent.
Either way, please post this as a proper patch on the mailing list.
Kevin
Re: [PATCH 1/2] rust: Add antoher variant for impl_vmstate_struct! macro
In reply to Paolo BonziniQuoted 8 lines · Paolo Bonzini
Il lun 4 ago 2025, 10:56 Martin Kletzander <[email protected]> ha scritto:
Quoted 6 lines
*BUT* of course I had to rebase the patches on top of current rust-next on Friday and there were some of your commits from Thursday which I now see actually fix all what I tried fixing before as well. I tried finding the previous commit on which I saw all the issues and after some rebuilding I could not. So it is now not even broken on rust-next.
Quoted 6 lines
Quoted 3 lines
This way I completely wasted your time, but at least learned something that's happening in the code. Sorry for that.
Uh no you didn't. It was broken.
But it is not now, neither master, nor rust-next, nor anything I tried from the reflog, which makes me suspicious about what I was developing this on. I distinctly remember the $vmsd.get() call in the macro which I presume was causing the copying due to VMStateDescription automatically implementing the Copy trait due to the bindgen invocation.
But that's nowhere to be found, including git log --walk-reflog and manually searching various history trees. And I doubted that you rewrote the history of the rust-next branch. Neither do I remember changing the macro until I found out that other in code changes did not help to fix it.
After some time I now managed to find it. It was the previous version of a commit 4cb0670e12c4, and that is nowhere to be found in rust-next at the moment, I guess fixes were incorporated while rebasing the branch on current master.
That was a wild ride, but I'm glad it all works (apart from the bash version indentation) on rust-next. I'll try to read up on what's next to help with, if anything.
Have a nice day,
Martin
Quoted 53 lines
Paolo
Martin
Quoted 50 lines
Quoted 8 lines
Paolo
rust/qemu-api/src/vmstate.rs| 11 +++++++++++rust/qemu-api/src/
Quoted 4 lines
rust/qemu-api/tests/vmstate_tests.rs | 1 + 2 files changed, 12 insertions(+)rust/qemu-api/src/ renamed from rust/qemu-api/src/vmstate.rs +0
-0
vmstate.rsQuoted 14 lines
Quoted 14 lines
index
b5c6b764fbba..716e52afe740100644rust/qemu-api/src/vmstate.rs +7
-0@@ -449,6 +449,17 @@ macro_rules! vmstate_validate {/// description of the struct. #[macro_export] macro_rules! impl_vmstate_struct { + ($type:ty, $(&)?$vmsd:ident) => { +unsafe impl $crate::vmstate::VMState for $type { +const BASE : $crate::bindings::VMStateField = + $crate::bindings::VMStateField { + vmsd: $vmsd.as_ref (), + size: ::core::mem::size_of ::<$type>(), + flags: $crate::bindings::VMStateFlags ::VMS_STRUCT ,Quoted 8 lines
Review
Quoted 8 lines
rust/qemu-api/src/vmstate.rs +4
-0+ ..$crate::zeroable::Zeroable ::ZERO + }; + } + }; ($type:ty, $vmsd:expr) => {unsafe impl $crate::vmstate::VMState for $type {const BASE : $crate::bindings::VMStateField = {(quoted diff) +0
-0b/rust/qemu-api/tests/
Quoted 10 lines
rust/qemu-api/tests/vmstate_tests.rs fn test_vmstate_struct_varray_uint8_wrapper() { · line 324
Quoted 10 lines
vmstate_tests.rs
index2c0670ba0eed..7d3180e6c2ea100644rust/qemu-api/tests/vmstate_tests.rs +1
-0@@ -320,6 +320,7 @@ fn test_vmstate_struct_varray_uint8_wrapper() {b"arr_a_wrap\0" );assert_eq! (foo_fields[5 ].num_offset,228 ); +assert_eq! (foo_fields[5 ].vmsd,VMSTATE_FOOA .as_ref ());assert! (unsafe { foo_fields[5 ].field_exists.unwrap ()(foo_b_p,0 )});
Quoted 6 lines
Quoted 6 lines
rust/qemu-api/tests/vmstate_tests.rs +0
-1// The last VMStateField in VMSTATE_FOOB. --2.50.1
Re: [PATCH 1/2] rust: Add antoher variant for impl_vmstate_struct! macro
Il lun 4 ago 2025, 14:04 Martin Kletzander <[email protected]> ha scritto:
Quoted 3 lines
But that's nowhere to be found, including git log --walk-reflog and manually searching various history trees. And I doubted that you rewrote the history of the rust-next branch
Yes, rust-next has its history rewritten.
By the way the bug was probably related to some compiler changes, because I am almost certain it worked a couple months ago.
Paolo
Re: [PATCH 2/2] tests/qemu-iotests: Indent expected error messages
In reply to Kevin WolfQuoted 38 lines · Kevin Wolf
Am 01.08.2025 um 21:09 hat Fabiano Rosas geschrieben:
Quoted 11 lines
Quoted 9 lines · Daniel P. Berrangé
Quoted 5 lines · Martin Kletzander
From: Martin Kletzander <[email protected]>
When running all tests the expected "killed" messages are indented differently than the actual ones, by three more spaces. Change it so that the messages match and tests pass.
This would break the tests on my system and CI too.
Quoted 7 lines
Quoted 3 lines
What distro are you seeing this on ?
I'm guessing this is a different in either valgrind or C library ?
It's bash, we have an open issue about it:
I see a patch has been posted to that bug, the most important part of which is this added filtering:
tests/qemu-iotests/common.filter 2025-07-25 11:39:22.419665788 +0000 renamed from tests/qemu-iotests/common.filter +1
-1@@ -75,7 +75,7 @@ _filter_qemu_io() _filter_win32 | \ gsed -e "s/[0-9]* ops\; [0-9/:. sec]* ([0-9/.inf]* [EPTGMKiBbytes]*\/sec and [0-9/.inf]* ops\/sec)/X ops\; XX:XX:XX.X (XXX YYY\/sec and XXX ops\/sec)/" \ -e "s/: line [0-9][0-9]*: *[0-9][0-9]*\( Aborted\| Killed\)/:\1/" \ - -e "s/qemu-io> //g" + -e "s/qemu-io> //g" -e '/Killed/{ s/ \{2,\}/ /}' } # replace occurrences of QEMU_PROG with "qemu"This approach makes sense to me, though I would have kept each sed expression on a separate line.
Quoted 7 lines
And given that the context line above includes "Aborted" as well, maybe have it here, too, though none of the actual test outputs have an Aborted message any more since commit 3f39447. Or we could have a cleanup patch first that removes the unused "Abort" above, just to keep things consistent.
Either way, please post this as a proper patch on the mailing list.
Just next try in tha attached patch with a fixed version of the sed command.
Werner
Signature · 2 lines
"Having a smoking section in a restaurant is like having
a peeing section in a swimming pool." -- Edward BurrFrom: Werner Fink <[email protected]>
Date: Fri, 08 Aug 06:41:23 +0000
Subject: [PATCH 1/1] Avoid dependency on padding on signal messages
New bash 5.3 uses a different padding for reporting job status.
Resolves: boo#1246830
--- tests/qemu-iotests/039.out | 10 +++++----- tests/qemu-iotests/061.out | 4 ++-- tests/qemu-iotests/137.out | 2 +- tests/qemu-iotests/common.filter | 2 +- 4 files changed, 9 insertions(+), 9 deletions(-)
tests/qemu-iotests/039.out +5 -5
@@ -11,7 +11,7 @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) incompatible_features [0] ERROR cluster 5 refcount=0 reference=1 ERROR OFLAG_COPIED data cluster: l2_entry=8000000000050000 refcount=0
@@ -46,7 +46,7 @@ read 512/512 bytes at offset 0 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) incompatible_features [0] ERROR cluster 5 refcount=0 reference=1 Rebuilding refcount structure
@@ -60,7 +60,7 @@ incompatible_features [] Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) incompatible_features [] No errors were found on the image.
@@ -79,7 +79,7 @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) incompatible_features [0] ERROR cluster 5 refcount=0 reference=1 ERROR OFLAG_COPIED data cluster: l2_entry=8000000000050000 refcount=0
@@ -89,7 +89,7 @@ Data may be corrupted, or further writes Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) incompatible_features [] No errors were found on the image. *** done
tests/qemu-iotests/061.out +2 -2
@@ -118,7 +118,7 @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 wrote 131072/131072 bytes at offset 0 128 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) magic 0x514649fb version 3 backing_file_offset 0x0
@@ -304,7 +304,7 @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 wrote 131072/131072 bytes at offset 0 128 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) magic 0x514649fb version 3 backing_file_offset 0x0
tests/qemu-iotests/137.out +1 -1
@@ -35,7 +35,7 @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGF qemu-io: Unsupported value 'blubb' for qcow2 option 'overlap-check'. Allowed are any of the following: none, constant, cached, all wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) +./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}" _qemu_proc_exec "${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" ) OK: Dirty bit not set Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 qemu-io: Parameter 'lazy-refcounts' expects 'on' or 'off'
tests/qemu-iotests/common.filter +1 -1
@@ -74,7 +74,7 @@ _filter_qemu_io() { _filter_win32 | \ gsed -e "s/[0-9]* ops\; [0-9/:. sec]* ([0-9/.inf]* [EPTGMKiBbytes]*\/sec and [0-9/.inf]* ops\/sec)/X ops\; XX:XX:XX.X (XXX YYY\/sec and XXX ops\/sec)/" \ - -e "s/: line [0-9][0-9]*: *[0-9][0-9]*\( Aborted\| Killed\)/:\1/" \ + -e "s/: line [0-9][0-9]*: *[0-9][0-9]*\( Aborted\| Killed\) \{2,\}/:\1 /" \ -e "s/qemu-io> //g" }
Martin Kletzander
Tested-byRe: [PATCH 2/2] tests/qemu-iotests: Indent expected error messages
Quoted 45 lines · Dr. Werner Fink
Quoted 44 lines · Kevin Wolf
Am 01.08.2025 um 21:09 hat Fabiano Rosas geschrieben:
Quoted 11 lines
Quoted 9 lines · Daniel P. Berrangé
Quoted 5 lines · Martin Kletzander
From: Martin Kletzander <[email protected]>
When running all tests the expected "killed" messages are indented differently than the actual ones, by three more spaces. Change it so that the messages match and tests pass.
This would break the tests on my system and CI too.
Quoted 7 lines
Quoted 3 lines
What distro are you seeing this on ?
I'm guessing this is a different in either valgrind or C library ?
It's bash, we have an open issue about it:
I see a patch has been posted to that bug, the most important part of which is this added filtering:
tests/qemu-iotests/common.filter 2025-07-25 11:39:22.419665788 +0000 renamed from tests/qemu-iotests/common.filter +1
-1@@ -75,7 +75,7 @@ _filter_qemu_io() _filter_win32 | \ gsed -e "s/[0-9]* ops\; [0-9/:. sec]* ([0-9/.inf]* [EPTGMKiBbytes]*\/sec and [0-9/.inf]* ops\/sec)/X ops\; XX:XX:XX.X (XXX YYY\/sec and XXX ops\/sec)/" \ -e "s/: line [0-9][0-9]*: *[0-9][0-9]*\( Aborted\| Killed\)/:\1/" \ - -e "s/qemu-io> //g" + -e "s/qemu-io> //g" -e '/Killed/{ s/ \{2,\}/ /}' } # replace occurrences of QEMU_PROG with "qemu"This approach makes sense to me, though I would have kept each sed expression on a separate line.
And given that the context line above includes "Aborted" as well, maybe have it here, too, though none of the actual test outputs have an Aborted message any more since commit 3f39447. Or we could have a cleanup patch first that removes the unused "Abort" above, just to keep things consistent.
Quoted 18 lines
Quoted 1 line
Either way, please post this as a proper patch on the mailing list.
Just next try in tha attached patch with a fixed version of the sed command.
Werner
Signature · 10 lines
"Having a smoking section in a restaurant is like having a peeing section in a swimming pool." -- Edward Burr From: Werner Fink <[email protected]> Date: Fri, 08 Aug 06:41:23 +0000 Subject: [PATCH 1/1] Avoid dependency on padding on signal messages New bash 5.3 uses a different padding for reporting job status. Resolves: boo#1246830
Not sure you meant that or
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/3050
Quoted 1 line
Signed-off-by Werner Fink <[email protected]>
but, FWIW, if that makes any difference,
Quoted 96 lines
---
tests/qemu-iotests/039.out | 10 +++++----- tests/qemu-iotests/061.out | 4 ++-- tests/qemu-iotests/137.out | 2 +- tests/qemu-iotests/common.filter | 2 +- 4 files changed, 9 insertions(+), 9 deletions(-)tests/qemu-iotests/039.out +0
-0@@ -11,7 +11,7 @@ No errors were found on the image.Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728
wrote 512/512 bytes at offset 0
512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}"_qemu_proc_exec"${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
+./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}"_qemu_proc_exec"${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )incompatible_features[0]
ERROR cluster 5 refcount=0 reference=1
ERROROFLAG_COPIEDdata cluster: l2_entry=8000000000050000 refcount=0tests/qemu-iotests/039.out +0
-0@@ -46,7 +46,7 @@ read 512/512 bytes at offset 0Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728
wrote 512/512 bytes at offset 0
512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}"_qemu_proc_exec"${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
+./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}"_qemu_proc_exec"${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )incompatible_features[0]
ERROR cluster 5 refcount=0 reference=1
Rebuilding refcount structuretests/qemu-iotests/039.out +0
-0@@ -60,7 +60,7 @@ incompatible_features []Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728
wrote 512/512 bytes at offset 0
512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}"_qemu_proc_exec"${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
+./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}"_qemu_proc_exec"${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )incompatible_features[]
No errors were found on the image.tests/qemu-iotests/039.out +0
-0@@ -79,7 +79,7 @@ No errors were found on the image.Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728
wrote 512/512 bytes at offset 0
512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}"_qemu_proc_exec"${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
+./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}"_qemu_proc_exec"${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )incompatible_features[0]
ERROR cluster 5 refcount=0 reference=1
ERROROFLAG_COPIEDdata cluster: l2_entry=8000000000050000 refcount=0tests/qemu-iotests/039.out +0
-0@@ -89,7 +89,7 @@ Data may be corrupted, or further writesFormatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728
wrote 512/512 bytes at offset 0
512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}"_qemu_proc_exec"${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
+./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}"_qemu_proc_exec"${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )incompatible_features[]
No errors were found on the image.
*** donetests/qemu-iotests/061.out +0
-0@@ -118,7 +118,7 @@ No errors were found on the image.Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
wrote 131072/131072 bytes at offset 0
128 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}"_qemu_proc_exec"${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
+./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}"_qemu_proc_exec"${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )magic 0x514649fb version 3 backing_file_offset 0x0tests/qemu-iotests/061.out +0
-0@@ -304,7 +304,7 @@ No errors were found on the image.Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
wrote 131072/131072 bytes at offset 0
128 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}"_qemu_proc_exec"${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
+./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}"_qemu_proc_exec"${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )magic 0x514649fb version 3 backing_file_offset 0x0tests/qemu-iotests/137.out +0
-0@@ -35,7 +35,7 @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFqemu-io: Unsupported value 'blubb' for qcow2 option 'overlap-check'. Allowed are any of the following: none, constant, cached, all wrote 512/512 bytes at offset 0
512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}"_qemu_proc_exec"${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
+./common.rc: Killed ( VALGRIND_QEMU="${VALGRIND_QEMU_IO}"_qemu_proc_exec"${VALGRIND_LOGFILE}" "$QEMU_IO_PROG" $QEMU_IO_ARGS "$@" )
OK: Dirty bit not set
Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
qemu-io: Parameter 'lazy-refcounts' expects 'on' or 'off'tests/qemu-iotests/common.filter +1
-1@@ -74,7 +74,7 @@ _filter_qemu_io() { _filter_win32 | \ gsed -e "s/[0-9]* ops\; [0-9/:. sec]* ([0-9/.inf]* [EPTGMKiBbytes]*\/sec and [0-9/.inf]* ops\/sec)/X ops\; XX:XX:XX.X (XXX YYY\/sec and XXX ops\/sec)/" \ - -e "s/: line [0-9][0-9]*: *[0-9][0-9]*\( Aborted\| Killed\)/:\1/" \ + -e "s/: line [0-9][0-9]*: *[0-9][0-9]*\( Aborted\| Killed\) \{2,\}/:\1 /" \ -e "s/qemu-io> //g" }
Kevin Wolf
AppliedRe: [PATCH 2/2] tests/qemu-iotests: Indent expected error messages
Am 11.08.2025 um 11:05 hat Martin Kletzander geschrieben:
Quoted 61 lines · Dr. Werner Fink
Quoted 60 lines · Kevin Wolf
Quoted 38 lines
Am 01.08.2025 um 21:09 hat Fabiano Rosas geschrieben:
Quoted 11 lines
Quoted 9 lines · Daniel P. Berrangé
Quoted 5 lines · Martin Kletzander
From: Martin Kletzander <[email protected]>
When running all tests the expected "killed" messages are indented differently than the actual ones, by three more spaces. Change it so that the messages match and tests pass.
This would break the tests on my system and CI too.
Quoted 7 lines
Quoted 3 lines
What distro are you seeing this on ?
I'm guessing this is a different in either valgrind or C library ?
It's bash, we have an open issue about it:
I see a patch has been posted to that bug, the most important part of which is this added filtering:
tests/qemu-iotests/common.filter 2025-07-25 11:39:22.419665788 +0000 renamed from tests/qemu-iotests/common.filter +1
-1@@ -75,7 +75,7 @@ _filter_qemu_io() _filter_win32 | \ gsed -e "s/[0-9]* ops\; [0-9/:. sec]* ([0-9/.inf]* [EPTGMKiBbytes]*\/sec and [0-9/.inf]* ops\/sec)/X ops\; XX:XX:XX.X (XXX YYY\/sec and XXX ops\/sec)/" \ -e "s/: line [0-9][0-9]*: *[0-9][0-9]*\( Aborted\| Killed\)/:\1/" \ - -e "s/qemu-io> //g" + -e "s/qemu-io> //g" -e '/Killed/{ s/ \{2,\}/ /}' } # replace occurrences of QEMU_PROG with "qemu"This approach makes sense to me, though I would have kept each sed expression on a separate line.
Quoted 7 lines
And given that the context line above includes "Aborted" as well, maybe have it here, too, though none of the actual test outputs have an Aborted message any more since commit 3f39447. Or we could have a cleanup patch first that removes the unused "Abort" above, just to keep things consistent.
Either way, please post this as a proper patch on the mailing list.
Just next try in tha attached patch with a fixed version of the sed command.
Werner
Signature · 6 lines
"Having a smoking section in a restaurant is like having a peeing section in a swimming pool." -- Edward Burr From: Werner Fink <[email protected]> Date: Fri, 08 Aug 06:41:23 +0000 Subject: [PATCH 1/1] Avoid dependency on padding on signal messages
Quoted 6 lines
Quoted 3 lines
New bash 5.3 uses a different padding for reporting job status.
Resolves: boo#1246830
Not sure you meant that or
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/3050
I just updated the commit message to contain both.
Quoted 5 lines
Quoted 1 line
Signed-off-by Werner Fink <[email protected]>but, FWIW, if that makes any difference,
Tested-by Martin Kletzander <[email protected]>
Thanks, applied to the block branch.
Kevin