Conversation
A legacy PCI interrupt is signalled level triggered and active low, while the IOAPIC defaults to the edge triggered, active high behavior of the ISA interrupts. A passed-through device therefore stops delivering interrupts after its first one, because the host only unmasks it once the guest has completed a level triggered interrupt. Correct the redirection entry where a legacy interrupt line is handed out.
![github-actions[bot]](https://avatars.githubusercontent.com/in/15368?s=60&v=4){kind=small}
There was a problem hiding this comment.
Benchmark Results
Details
| Benchmark | Current: 6bdc928 | Previous: 2e23902 | Performance Ratio |
|---|---|---|---|
| startup_benchmark Build Time | 82.00 s |
80.34 s |
1.02 ❗ |
| startup_benchmark File Size | 0.76 MB |
0.80 MB |
0.96 ❗ |
| Startup Time - 1 core | 0.77 s (±0.03 s) |
0.75 s (±0.02 s) |
1.03 |
| Startup Time - 2 cores | 0.79 s (±0.02 s) |
0.74 s (±0.02 s) |
1.07 ❗ |
| Startup Time - 4 cores | 0.77 s (±0.02 s) |
0.74 s (±0.02 s) |
1.04 |
| multithreaded_benchmark Build Time | 83.99 s |
82.11 s |
1.02 ❗ |
| multithreaded_benchmark File Size | 0.88 MB |
0.86 MB |
1.02 ❗ |
| Multithreaded Pi Efficiency - 2 Threads | 82.82 % (±7.61 %) |
85.89 % (±6.61 %) |
0.96 |
| Multithreaded Pi Efficiency - 4 Threads | 41.80 % (±3.27 %) |
43.43 % (±2.56 %) |
0.96 |
| Multithreaded Pi Efficiency - 8 Threads | 24.36 % (±1.67 %) |
25.76 % (±1.53 %) |
0.95 |
| micro_benchmarks Build Time | 83.30 s |
80.40 s |
1.04 ❗ |
| micro_benchmarks File Size | 0.88 MB |
0.86 MB |
1.02 ❗ |
| Scheduling time - 1 thread | 65.34 ticks (±3.12 ticks) |
62.65 ticks (±4.06 ticks) |
1.04 |
| Scheduling time - 2 threads | 36.50 ticks (±4.46 ticks) |
34.08 ticks (±4.10 ticks) |
1.07 |
| Micro - Time for syscall (getpid) | 4.59 ticks (±0.79 ticks) |
3.45 ticks (±0.58 ticks) |
1.33 ❗ |
| Memcpy speed - (built_in) block size 4096 | 82934.28 MByte/s (±57293.20 MByte/s) |
82448.38 MByte/s (±56997.13 MByte/s) |
1.01 |
| Memcpy speed - (built_in) block size 1048576 | 30283.60 MByte/s (±24429.94 MByte/s) |
30585.98 MByte/s (±24707.84 MByte/s) |
0.99 |
| Memcpy speed - (built_in) block size 16777216 | 27117.81 MByte/s (±22407.18 MByte/s) |
26340.06 MByte/s (±21720.96 MByte/s) |
1.03 |
| Memset speed - (built_in) block size 4096 | 83321.19 MByte/s (±57554.04 MByte/s) |
82292.76 MByte/s (±56891.50 MByte/s) |
1.01 |
| Memset speed - (built_in) block size 1048576 | 30980.06 MByte/s (±24852.16 MByte/s) |
31323.85 MByte/s (±25145.86 MByte/s) |
0.99 |
| Memset speed - (built_in) block size 16777216 | 27890.83 MByte/s (±22889.81 MByte/s) |
27104.68 MByte/s (±22209.94 MByte/s) |
1.03 |
| Memcpy speed - (rust) block size 4096 | 76226.44 MByte/s (±52907.32 MByte/s) |
74097.96 MByte/s (±51811.44 MByte/s) |
1.03 |
| Memcpy speed - (rust) block size 1048576 | 30287.69 MByte/s (±24527.98 MByte/s) |
30361.60 MByte/s (±24602.37 MByte/s) |
1.00 |
| Memcpy speed - (rust) block size 16777216 | 26737.86 MByte/s (±22175.56 MByte/s) |
27625.34 MByte/s (±22806.88 MByte/s) |
0.97 |
| Memset speed - (rust) block size 4096 | 76735.85 MByte/s (±53273.78 MByte/s) |
74373.47 MByte/s (±51976.48 MByte/s) |
1.03 |
| Memset speed - (rust) block size 1048576 | 31024.39 MByte/s (±24953.38 MByte/s) |
31110.89 MByte/s (±25033.24 MByte/s) |
1.00 |
| Memset speed - (rust) block size 16777216 | 27504.63 MByte/s (±22657.11 MByte/s) |
28386.93 MByte/s (±23265.03 MByte/s) |
0.97 |
| alloc_benchmarks Build Time | 76.92 s |
74.76 s |
1.03 ❗ |
| alloc_benchmarks File Size | 0.84 MB |
0.87 MB |
0.96 ❗ |
| Allocations - Allocation success | 91.34 % |
91.31 % |
1.00 ❗ |
| Allocations - Deallocation success | 100.00 % |
100.00 % |
1 |
| Allocations - Pre-fail Allocations | 61.54 % |
61.44 % |
1.00 ❗ |
| Allocations - Average Allocation time | 2525.89 Ticks (±778.41 Ticks) |
5860.58 Ticks (±98.43 Ticks) |
0.43 ❗ |
| Allocations - Average Allocation time (no fail) | 3560.40 Ticks (±673.34 Ticks) |
6554.81 Ticks (±92.86 Ticks) |
0.54 ❗ |
| Allocations - Average Deallocation time | 683.68 Ticks (±99.02 Ticks) |
1805.01 Ticks (±250.35 Ticks) |
0.38 ❗ |
| mutex_benchmark Build Time | 81.21 s |
79.82 s |
1.02 ❗ |
| mutex_benchmark File Size | 0.88 MB |
0.86 MB |
1.02 ❗ |
| Mutex Stress Test Average Time per Iteration - 1 Threads | 12.10 ns (±0.30 ns) |
12.10 ns (±0.41 ns) |
1 |
| Mutex Stress Test Average Time per Iteration - 2 Threads | 40.50 ns (±1.96 ns) |
40.26 ns (±1.68 ns) |
1.01 |
This comment was automatically generated by workflow using github-action-benchmark.
Determining the size of a BAR writes all-ones into it and restores the
previous value afterwards. While decoding is enabled, the device decodes
those intermediate values as real addresses and relocates itself. For a
passed-through device the host follows that relocation and tries to map
the BAR outside of the guest, which aborts the VM:
kvm_set_user_memory_region: KVM_SET_USER_MEMORY_REGION failed,
slot=5, start=0xfffffffffe000000, size=0x400000: Invalid argument
Disable decoding around the probe, as the PCI specification requires,
and route the informational bus dump through get_bar so that it is
covered as well.
The initialization sequence follows the ixy.rs user space driver and the X550 datasheet. The driver operates a single receive and transmit queue. It prefers MSI-X over the legacy interrupt, because a hypervisor is not obliged to offer one for a passed-through device and QEMU in fact does not, which leaves MSI-X as the only way to receive interrupts at all. Should neither be available, the device still makes progress when the kernel is built with the idle-poll feature. This makes the msix module available outside of virtio.
The initialization sequence follows the ixy.rs user space driver and the X550 datasheet. The driver operates a single receive and transmit queue. It prefers MSI-X over the legacy interrupt, because a hypervisor is not obliged to offer one for a passed-through device and QEMU in fact does not, which leaves MSI-X as the only way to receive interrupts at all. Should neither be available, the device still makes progress when the kernel is built with the idle-poll feature. This makes the msix module available outside of virtio.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The initialization sequence follows the ixy.rs user space driver and the X550 datasheet. The driver operates a single receive and transmit queue.