commit afc545da381ba0c651b2658966ac737032676f01 upstream.
This patch fixes an issue in the function xenbus_dev_probe(). In the
xenbus_dev_probe() function, within the if (err) branch at line 313, the
program incorrectly returns err directly without releasing the resources
allocated by err = drv->probe(dev, id). As the return value is non-zero,
the upper layers assume the processing logic has failed. However, the probe
operation was performed earlier without a corresponding remove operation.
Since the probe actually allocates resources, failing to perform the remove
operation could lead to problems.
To fix this issue, we followed the resource release logic of the
xenbus_dev_remove() function by adding a new block fail_remove before the
fail_put block. After entering the branch if (err) at line 313, the
function will use a goto statement to jump to the fail_remove block,
ensuring that the previously acquired resources are correctly released,
thus preventing the reference count leak.
This bug was identified by an experimental static analysis tool developed
by our team. The tool specializes in analyzing reference count operations
and detecting potential issues where resources are not properly managed.
In this case, the tool flagged the missing release operation as a
potential problem, which led to the development of this patch.
Fixes: 4bac07c993d0 ("xen: add the Xenbus sysfs and virtual device hotplug driver")
Cc: stable@vger.kernel.org
Signed-off-by: Qiu-ji Chen <chenqiuji666@gmail.com>
Reviewed-by: Juergen Gross <jgross@suse.com>
Message-ID: <20241105130919.4621-1-chenqiuji666@gmail.com>
Signed-off-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 9f40ec84a7976d95c34e7cc070939deb103652b0 ]
When checking a memory buffer to be consecutive in machine memory,
the alignment needs to be checked, too. Failing to do so might result
in DMA memory not being aligned according to its requested size,
leading to error messages like:
4xxx 0000:2b:00.0: enabling device (0140 -> 0142)
4xxx 0000:2b:00.0: Ring address not aligned
4xxx 0000:2b:00.0: Failed to initialise service qat_crypto
4xxx 0000:2b:00.0: Resetting device qat_dev0
4xxx: probe of 0000:2b:00.0 failed with error -14
Fixes: 9435cce87950 ("xen/swiotlb: Add support for 64KB page granularity")
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
Signed-off-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit fa765c4b4aed2d64266b694520ecb025c862c5a9 upstream.
shutdown_pirq and startup_pirq are not taking the
irq_mapping_update_lock because they can't due to lock inversion. Both
are called with the irq_desc->lock being taking. The lock order,
however, is first irq_mapping_update_lock and then irq_desc->lock.
This opens multiple races:
- shutdown_pirq can be interrupted by a function that allocates an event
channel:
CPU0 CPU1
shutdown_pirq {
xen_evtchn_close(e)
__startup_pirq {
EVTCHNOP_bind_pirq
-> returns just freed evtchn e
set_evtchn_to_irq(e, irq)
}
xen_irq_info_cleanup() {
set_evtchn_to_irq(e, -1)
}
}
Assume here event channel e refers here to the same event channel
number.
After this race the evtchn_to_irq mapping for e is invalid (-1).
- __startup_pirq races with __unbind_from_irq in a similar way. Because
__startup_pirq doesn't take irq_mapping_update_lock it can grab the
evtchn that __unbind_from_irq is currently freeing and cleaning up. In
this case even though the event channel is allocated, its mapping can
be unset in evtchn_to_irq.
The fix is to first cleanup the mappings and then close the event
channel. In this way, when an event channel gets allocated it's
potential previous evtchn_to_irq mappings are guaranteed to be unset already.
This is also the reverse order of the allocation where first the event
channel is allocated and then the mappings are setup.
On a 5.10 kernel prior to commit 3fcdaf3d7634 ("xen/events: modify internal
[un]bind interfaces"), we hit a BUG like the following during probing of NVMe
devices. The issue is that during nvme_setup_io_queues, pci_free_irq
is called for every device which results in a call to shutdown_pirq.
With many nvme devices it's therefore likely to hit this race during
boot because there will be multiple calls to shutdown_pirq and
startup_pirq are running potentially in parallel.
------------[ cut here ]------------
blkfront: xvda: barrier or flush: disabled; persistent grants: enabled; indirect descriptors: enabled; bounce buffer: enabled
kernel BUG at drivers/xen/events/events_base.c:499!
invalid opcode: 0000 [#1] SMP PTI
CPU: 44 PID: 375 Comm: kworker/u257:23 Not tainted 5.10.201-191.748.amzn2.x86_64 #1
Hardware name: Xen HVM domU, BIOS 4.11.amazon 08/24/2006
Workqueue: nvme-reset-wq nvme_reset_work
RIP: 0010:bind_evtchn_to_cpu+0xdf/0xf0
Code: 5d 41 5e c3 cc cc cc cc 44 89 f7 e8 2b 55 ad ff 49 89 c5 48 85 c0 0f 84 64 ff ff ff 4c 8b 68 30 41 83 fe ff 0f 85 60 ff ff ff <0f> 0b 66 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 0f 1f 44 00 00
RSP: 0000:ffffc9000d533b08 EFLAGS: 00010046
RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000006
RDX: 0000000000000028 RSI: 00000000ffffffff RDI: 00000000ffffffff
RBP: ffff888107419680 R08: 0000000000000000 R09: ffffffff82d72b00
R10: 0000000000000000 R11: 0000000000000000 R12: 00000000000001ed
R13: 0000000000000000 R14: 00000000ffffffff R15: 0000000000000002
FS: 0000000000000000(0000) GS:ffff88bc8b500000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000000 CR3: 0000000002610001 CR4: 00000000001706e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
? show_trace_log_lvl+0x1c1/0x2d9
? show_trace_log_lvl+0x1c1/0x2d9
? set_affinity_irq+0xdc/0x1c0
? __die_body.cold+0x8/0xd
? die+0x2b/0x50
? do_trap+0x90/0x110
? bind_evtchn_to_cpu+0xdf/0xf0
? do_error_trap+0x65/0x80
? bind_evtchn_to_cpu+0xdf/0xf0
? exc_invalid_op+0x4e/0x70
? bind_evtchn_to_cpu+0xdf/0xf0
? asm_exc_invalid_op+0x12/0x20
? bind_evtchn_to_cpu+0xdf/0xf0
? bind_evtchn_to_cpu+0xc5/0xf0
set_affinity_irq+0xdc/0x1c0
irq_do_set_affinity+0x1d7/0x1f0
irq_setup_affinity+0xd6/0x1a0
irq_startup+0x8a/0xf0
__setup_irq+0x639/0x6d0
? nvme_suspend+0x150/0x150
request_threaded_irq+0x10c/0x180
? nvme_suspend+0x150/0x150
pci_request_irq+0xa8/0xf0
? __blk_mq_free_request+0x74/0xa0
queue_request_irq+0x6f/0x80
nvme_create_queue+0x1af/0x200
nvme_create_io_queues+0xbd/0xf0
nvme_setup_io_queues+0x246/0x320
? nvme_irq_check+0x30/0x30
nvme_reset_work+0x1c8/0x400
process_one_work+0x1b0/0x350
worker_thread+0x49/0x310
? process_one_work+0x350/0x350
kthread+0x11b/0x140
? __kthread_bind_mask+0x60/0x60
ret_from_fork+0x22/0x30
Modules linked in:
---[ end trace a11715de1eee1873 ]---
Fixes: d46a78b05c0e ("xen: implement pirq type event channels")
Cc: stable@vger.kernel.org
Co-debugged-by: Andrew Panyakin <apanyaki@amazon.com>
Signed-off-by: Maximilian Heyne <mheyne@amazon.de>
Reviewed-by: Juergen Gross <jgross@suse.com>
Link: https://lore.kernel.org/r/20240124163130.31324-1-mheyne@amazon.de
Signed-off-by: Juergen Gross <jgross@suse.com>
[apanyaki: backport to v5.10-stable]
Signed-off-by: Andrew Paniakin <apanyaki@amazon.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 2d2db7d40254d5fb53b11ebd703cd1ed0c5de7a1 ]
DO NOT access the underlying struct page of an sg table exported
by DMA-buf in dmabuf_imp_to_refs(), this is not allowed.
Please see drivers/dma-buf/dma-buf.c:mangle_sg_table() for details.
Fortunately, here (for special Xen device) we can avoid using
pages and calculate gfns directly from dma addresses provided by
the sg table.
Suggested-by: Daniel Vetter <daniel@ffwll.ch>
Signed-off-by: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
Acked-by: Daniel Vetter <daniel@ffwll.ch>
Link: https://lore.kernel.org/r/20240107103426.2038075-1-olekstysh@gmail.com
Signed-off-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit bff2a2d453a1b683378b4508b86b84389f551a00 upstream.
There's a bug that when using the XEN hypervisor with bios with large
multi-page bio vectors on NVMe, the kernel deadlocks [1].
The deadlocks are caused by inability to map a large bio vector -
dma_map_sgtable always returns an error, this gets propagated to the block
layer as BLK_STS_RESOURCE and the block layer retries the request
indefinitely.
XEN uses the swiotlb framework to map discontiguous pages into contiguous
runs that are submitted to the PCIe device. The swiotlb framework has a
limitation on the length of a mapping - this needs to be announced with
the max_mapping_size method to make sure that the hardware drivers do not
create larger mappings.
Without max_mapping_size, the NVMe block driver would create large
mappings that overrun the maximum mapping size.
Reported-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
Link: https://lore.kernel.org/stable/ZTNH0qtmint%2FzLJZ@mail-itl/ [1]
Tested-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
Suggested-by: Christoph Hellwig <hch@lst.de>
Cc: stable@vger.kernel.org
Signed-off-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Acked-by: Stefano Stabellini <sstabellini@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/151bef41-e817-aea9-675-a35fdac4ed@redhat.com
Signed-off-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 47d970204054f859f35a2237baa75c2d84fcf436 ]
When delaying eoi handling of events, the related elements are queued
into the percpu lateeoi list. In case the list isn't empty, the
elements should be sorted by the time when eoi handling is to happen.
Unfortunately a new element will never be queued at the start of the
list, even if it has a handling time lower than all other list
elements.
Fix that by handling that case the same way as for an empty list.
Fixes: e99502f76271 ("xen/events: defer eoi in case of excessive number of events")
Reported-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Oleksandr Tyshchenko <oleksandr_tyshchenko@epam.com>
Signed-off-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2c269f42d0f382743ab230308b836ffe5ae9b2ae ]
Linux enables MSI-X before disabling INTx, but keeps MSI-X masked until
the table is filled. Then it disables INTx just before clearing MASKALL
bit. Currently this approach is rejected by xen-pciback.
According to the PCIe spec, device cannot use INTx when MSI/MSI-X is
enabled (in other words: enabling MSI/MSI-X implicitly disables INTx).
Change the logic to consider INTx disabled if MSI/MSI-X is enabled. This
applies to three places:
- checking currently enabled interrupts type,
- transition to MSI/MSI-X - where INTx would be implicitly disabled,
- clearing INTx disable bit - which can be allowed even if MSI/MSI-X is
enabled, as device should consider INTx disabled anyway in that case
Fixes: 5e29500eba2a ("xen-pciback: Allow setting PCI_MSIX_FLAGS_MASKALL too")
Signed-off-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
Acked-by: Juergen Gross <jgross@suse.com>
Link: https://lore.kernel.org/r/20231016131348.1734721-1-marmarek@invisiblethingslab.com
Signed-off-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
