[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Stable-8.2.3 21/87] hw/nvme: add machine compatibility parameter to ena
From: |
Michael Tokarev |
Subject: |
[Stable-8.2.3 21/87] hw/nvme: add machine compatibility parameter to enable msix exclusive bar |
Date: |
Wed, 10 Apr 2024 10:21:54 +0300 |
From: Klaus Jensen <k.jensen@samsung.com>
Commit 1901b4967c3f ("hw/block/nvme: move msix table and pba to BAR 0")
moved the MSI-X table and PBA to BAR 0 to make room for enabling CMR and
PMR at the same time. As reported by Julien Grall in #2184, this breaks
migration through system hibernation.
Add a machine compatibility parameter and set it on machines pre 6.0 to
enable the old behavior automatically, restoring the hibernation
migration support.
Cc: qemu-stable@nongnu.org
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2184
Fixes: 1901b4967c3f ("hw/block/nvme: move msix table and pba to BAR 0")
Reported-by: Julien Grall julien@xen.org
Tested-by: Julien Grall julien@xen.org
Reviewed-by: Jesper Wendel Devantier <foss@defmacro.it>
Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
(cherry picked from commit fa905f65c5549703279f68c253914799b10ada47)
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
diff --git a/hw/core/machine.c b/hw/core/machine.c
index 0c17398141..3c08a894fb 100644
--- a/hw/core/machine.c
+++ b/hw/core/machine.c
@@ -97,6 +97,7 @@ GlobalProperty hw_compat_5_2[] = {
{ "PIIX4_PM", "smm-compat", "on"},
{ "virtio-blk-device", "report-discard-granularity", "off" },
{ "virtio-net-pci-base", "vectors", "3"},
+ { "nvme", "msix-exclusive-bar", "on"},
};
const size_t hw_compat_5_2_len = G_N_ELEMENTS(hw_compat_5_2);
diff --git a/hw/nvme/ctrl.c b/hw/nvme/ctrl.c
index 74da28a35a..585bd3b397 100644
--- a/hw/nvme/ctrl.c
+++ b/hw/nvme/ctrl.c
@@ -7798,6 +7798,11 @@ static bool nvme_check_params(NvmeCtrl *n, Error **errp)
}
if (n->pmr.dev) {
+ if (params->msix_exclusive_bar) {
+ error_setg(errp, "not enough BARs available to enable PMR");
+ return false;
+ }
+
if (host_memory_backend_is_mapped(n->pmr.dev)) {
error_setg(errp, "can't use already busy memdev: %s",
object_get_canonical_path_component(OBJECT(n->pmr.dev)));
@@ -8101,24 +8106,38 @@ static bool nvme_init_pci(NvmeCtrl *n, PCIDevice
*pci_dev, Error **errp)
pcie_ari_init(pci_dev, 0x100);
}
- /* add one to max_ioqpairs to account for the admin queue pair */
- bar_size = nvme_mbar_size(n->params.max_ioqpairs + 1, n->params.msix_qsize,
- &msix_table_offset, &msix_pba_offset);
+ if (n->params.msix_exclusive_bar && !pci_is_vf(pci_dev)) {
+ bar_size = nvme_mbar_size(n->params.max_ioqpairs + 1, 0, NULL, NULL);
+ memory_region_init_io(&n->iomem, OBJECT(n), &nvme_mmio_ops, n, "nvme",
+ bar_size);
+ pci_register_bar(pci_dev, 0, PCI_BASE_ADDRESS_SPACE_MEMORY |
+ PCI_BASE_ADDRESS_MEM_TYPE_64, &n->iomem);
+ ret = msix_init_exclusive_bar(pci_dev, n->params.msix_qsize, 4, errp);
+ } else {
+ assert(n->params.msix_qsize >= 1);
- memory_region_init(&n->bar0, OBJECT(n), "nvme-bar0", bar_size);
- memory_region_init_io(&n->iomem, OBJECT(n), &nvme_mmio_ops, n, "nvme",
- msix_table_offset);
- memory_region_add_subregion(&n->bar0, 0, &n->iomem);
+ /* add one to max_ioqpairs to account for the admin queue pair */
+ bar_size = nvme_mbar_size(n->params.max_ioqpairs + 1,
+ n->params.msix_qsize, &msix_table_offset,
+ &msix_pba_offset);
- if (pci_is_vf(pci_dev)) {
- pcie_sriov_vf_register_bar(pci_dev, 0, &n->bar0);
- } else {
- pci_register_bar(pci_dev, 0, PCI_BASE_ADDRESS_SPACE_MEMORY |
- PCI_BASE_ADDRESS_MEM_TYPE_64, &n->bar0);
+ memory_region_init(&n->bar0, OBJECT(n), "nvme-bar0", bar_size);
+ memory_region_init_io(&n->iomem, OBJECT(n), &nvme_mmio_ops, n, "nvme",
+ msix_table_offset);
+ memory_region_add_subregion(&n->bar0, 0, &n->iomem);
+
+ if (pci_is_vf(pci_dev)) {
+ pcie_sriov_vf_register_bar(pci_dev, 0, &n->bar0);
+ } else {
+ pci_register_bar(pci_dev, 0, PCI_BASE_ADDRESS_SPACE_MEMORY |
+ PCI_BASE_ADDRESS_MEM_TYPE_64, &n->bar0);
+ }
+
+ ret = msix_init(pci_dev, n->params.msix_qsize,
+ &n->bar0, 0, msix_table_offset,
+ &n->bar0, 0, msix_pba_offset, 0, errp);
}
- ret = msix_init(pci_dev, n->params.msix_qsize,
- &n->bar0, 0, msix_table_offset,
- &n->bar0, 0, msix_pba_offset, 0, errp);
+
if (ret == -ENOTSUP) {
/* report that msix is not supported, but do not error out */
warn_report_err(*errp);
@@ -8422,6 +8441,8 @@ static Property nvme_props[] = {
params.sriov_max_vi_per_vf, 0),
DEFINE_PROP_UINT8("sriov_max_vq_per_vf", NvmeCtrl,
params.sriov_max_vq_per_vf, 0),
+ DEFINE_PROP_BOOL("msix-exclusive-bar", NvmeCtrl, params.msix_exclusive_bar,
+ false),
DEFINE_PROP_END_OF_LIST(),
};
diff --git a/hw/nvme/nvme.h b/hw/nvme/nvme.h
index 5f2ae7b28b..eccf14acc9 100644
--- a/hw/nvme/nvme.h
+++ b/hw/nvme/nvme.h
@@ -522,6 +522,7 @@ typedef struct NvmeParams {
uint16_t sriov_vi_flexible;
uint8_t sriov_max_vq_per_vf;
uint8_t sriov_max_vi_per_vf;
+ bool msix_exclusive_bar;
} NvmeParams;
typedef struct NvmeCtrl {
--
2.39.2
- [Stable-8.2.3 15/87] e1000e: fix link state on resume, (continued)
- [Stable-8.2.3 15/87] e1000e: fix link state on resume, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 12/87] make-release: switch to .xz format by default, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 14/87] igb: fix link state on resume, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 16/87] Avoid unaligned fetch in ladr_match(), Michael Tokarev, 2024/04/10
- [Stable-8.2.3 19/87] hw/nvme: fix invalid check on mcl, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 17/87] xen: Drop out of coroutine context xen_invalidate_map_cache_entry, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 20/87] hw/nvme: generalize the mbar size helper, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 18/87] hw/nvme: separate 'serial' property for VFs, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 22/87] hw/nvme: Use pcie_sriov_num_vfs(), Michael Tokarev, 2024/04/10
- [Stable-8.2.3 25/87] hw/audio/virtio-sound: return correct command response size, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 21/87] hw/nvme: add machine compatibility parameter to enable msix exclusive bar,
Michael Tokarev <=
- [Stable-8.2.3 23/87] pcie_sriov: Validate NumVFs, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 24/87] hmat acpi: Fix out of bounds access due to missing use of indirection, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 26/87] migration: Skip only empty block devices, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 30/87] nbd/server: Fix race in draining the export, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 29/87] nbd/server: introduce NBDClient->lock to protect fields, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 31/87] iotests: Add test for reset/AioContext switches with NBD exports, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 27/87] mirror: Don't call job_pause_point() under graph lock, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 28/87] nbd/server: only traverse NBDExport->clients from main loop thread, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 37/87] target/i386: fix direction of "32-bit MMU" test, Michael Tokarev, 2024/04/10
- [Stable-8.2.3 32/87] tests/unit: Bump test-aio-multithread test timeout to 2 minutes, Michael Tokarev, 2024/04/10