From: Davidlohr Bueso <dave@stgolabs.net>
To: Keith Busch <kbusch@meta.com>
Cc: linux-pci@vger.kernel.org, bhelgaas@google.com,
Keith Busch <kbusch@kernel.org>,
Jonathan Cameron <Jonathan.Cameron@huawei.com>
Subject: Re: [PATCHv2 2/5] pci: make pci_destroy_dev concurrent safe
Date: Wed, 2 Oct 2024 19:34:13 -0700 [thread overview]
Message-ID: <20241003023354.txfw7w4ud247h5va@offworld> (raw)
In-Reply-To: <20240827192826.710031-3-kbusch@meta.com>
On Tue, 27 Aug 2024, Keith Busch wrote:
>+static inline bool pci_dev_test_and_set_removed(struct pci_dev *dev)
>+{
>+ return test_and_set_bit(PCI_DEV_REMOVED, &dev->priv_flags);
>+}
Same ordering/dependency description observations as mentioned in
patch 1 (both these cases are fully ordered).
>+
> #ifdef CONFIG_PCIEAER
> #include <linux/aer.h>
>
>diff --git a/drivers/pci/remove.c b/drivers/pci/remove.c
>index ec3064a115bf8..8284ab20949c9 100644
>--- a/drivers/pci/remove.c
>+++ b/drivers/pci/remove.c
>@@ -29,7 +29,7 @@ static void pci_stop_dev(struct pci_dev *dev)
>
> static void pci_destroy_dev(struct pci_dev *dev)
> {
>- if (!dev->dev.kobj.parent)
>+ if (pci_dev_test_and_set_removed(dev))
Doesn't this want to be if (!pci_dev_test_and_set_removed()) ?
This also fixes a splat when triggering a removal when you add
subordinate refcounting is added:
https://git.kernel.org/pub/scm/linux/kernel/git/kbusch/linux.git/commit/?h=pci-bus-locking-2024-09-09&id=3883c485d5e45b5e17f685f77ff4020bec162336
fyi:
[ 22.739614] BUG: kernel NULL pointer dereference, address: 0000000000000028
[ 22.739910] #PF: supervisor read access in kernel mode
[ 22.740132] #PF: error_code(0x0000) - not-present page
[ 22.740351] PGD 0 P4D 0
[ 22.740468] Oops: Oops: 0000 [#1] PREEMPT SMP KASAN PTI
[ 22.740695] CPU: 0 UID: 0 PID: 266 Comm: bash Tainted: G B 6.11.0-rc1-g3883c485d5e4-dirty #13
[ 22.741111] Tainted: [B]=BAD_PAGE
[ 22.741258] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.16.3-0-ga6ed6b701f0a-prebuilt.qemu.org 04/01/2014
[ 22.741727] RIP: 0010:pcie_aspm_check_latency.isra.0+0x192/0x4d0
[ 22.741990] Code: 18 e8 e2 6f 6f ff 48 89 df e8 aa ad 92 ff 4c 8b 2b 49 8d 7d 18 e8 9e ad 92 ff 49 8b 6d 18 4c 8d 65 28 4c f
[ 22.743438] RSP: 0018:ffff88800554f970 EFLAGS: 00010282
[ 22.743673] RAX: 0000000000000001 RBX: ffff888001cc0c80 RCX: 0000000000000001
[ 22.743976] RDX: ffff88804c63ce00 RSI: 0000000000000000 RDI: 0000000000000007
[ 22.744285] RBP: 0000000000000000 R08: 0000000000000001 R09: fffffbfff7b7275c
[ 22.744596] R10: 0000000000000000 R11: 0000000000000001 R12: 0000000000000028
[ 22.744906] R13: ffff888001f62000 R14: 0000000000000040 R15: 00000000000003e8
[ 22.745216] FS: 00007f14a687f740(0000) GS:ffff88806d200000(0000) knlGS:0000000000000000
[ 22.745565] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 22.745823] CR2: 0000000000000028 CR3: 000000004dc06000 CR4: 00000000000006f0
[ 22.746132] Call Trace:
[ 22.746246] <TASK>
[ 22.746346] ? show_regs+0x8c/0xa0
[ 22.746507] ? __die+0x2c/0x80
[ 22.746652] ? page_fault_oops+0x31a/0x830
[ 22.746839] ? __pfx_page_fault_oops+0x10/0x10
[ 22.747042] ? is_prefetch.constprop.0+0x9b/0x450
[ 22.747253] ? pcie_aspm_check_latency.isra.0+0x192/0x4d0
[ 22.747495] ? __pfx_is_prefetch.constprop.0+0x10/0x10
[ 22.747725] ? pcie_aspm_check_latency.isra.0+0x192/0x4d0
[ 22.747966] ? search_module_extables+0x93/0xc0
[ 22.748173] ? fixup_exception+0xd7/0x560
[ 22.748358] ? kernelmode_fixup_or_oops.constprop.0+0x9c/0xc0
[ 22.748613] ? __bad_area_nosemaphore+0x2f8/0x420
[ 22.748826] ? lock_mm_and_find_vma+0x90/0x4f0
[ 22.749028] ? do_user_addr_fault+0x58a/0xc80
[ 22.749226] ? rcu_is_watching+0x20/0x50
[ 22.749407] ? exc_page_fault+0x5c/0xd0
[ 22.749586] ? asm_exc_page_fault+0x26/0x30
[ 22.749776] ? pcie_aspm_check_latency.isra.0+0x192/0x4d0
[ 22.750020] ? __pfx_pcie_aspm_check_latency.isra.0+0x10/0x10
[ 22.750278] ? mark_held_locks+0x65/0x90
[ 22.750457] ? kobject_get+0x95/0x110
[ 22.750629] pcie_update_aspm_capable+0x128/0x1c0
[ 22.750843] pcie_aspm_exit_link_state+0x137/0x1e0
[ 22.751059] pci_remove_bus_device+0x15b/0x200
[ 22.751260] pci_remove_bus+0x4a/0x130
[ 22.751432] pci_remove_bus_device+0x88/0x200
[ 22.751631] pci_remove_bus+0x4a/0x130
[ 22.751802] pci_remove_bus_device+0x88/0x200
[ 22.752000] pci_remove_bus+0x4a/0x130
[ 22.752172] pci_remove_bus_device+0x88/0x200
[ 22.752370] pci_stop_and_remove_bus_device_locked+0x22/0x30
[ 22.752622] remove_store+0x125/0x140
[ 22.752791] ? __pfx_remove_store+0x10/0x10
[ 22.752981] ? __pfx___mutex_lock+0x10/0x10
[ 22.753170] ? __pfx__copy_from_iter+0x10/0x10
[ 22.753372] ? __pfx_remove_store+0x10/0x10
[ 22.753562] dev_attr_store+0x46/0x70
[ 22.753752] ? __pfx_dev_attr_store+0x10/0x10
[ 22.753965] sysfs_kf_write+0xa0/0xc0
[ 22.754142] kernfs_fop_write_iter+0x23d/0x300
[ 22.754346] ? __pfx_sysfs_kf_write+0x10/0x10
[ 22.754549] vfs_write+0x508/0xa90
[ 22.754709] ? __pfx_kernfs_fop_write_iter+0x10/0x10
[ 22.754935] ? __pfx_vfs_write+0x10/0x10
[ 22.755118] ? __fget_light+0xcd/0x120
[ 22.755295] ksys_write+0x108/0x200
[ 22.755458] ? __pfx_ksys_write+0x10/0x10
[ 22.755644] ? mark_held_locks+0x24/0x90
[ 22.755826] do_syscall_64+0xc1/0x1d0
[ 22.755998] entry_SYSCALL_64_after_hwframe+0x77/0x7f
[ 22.756229] RIP: 0033:0x7f14a697a240
[ 22.756394] Code: 40 00 48 8b 15 c1 9b 0d 00 f7 d8 64 89 02 48 c7 c0 ff ff ff ff eb b7 0f 1f 00 80 3d a1 23 0e 00 00 74 17 9
[ 22.757184] RSP: 002b:00007ffdc41b64e8 EFLAGS: 00000202 ORIG_RAX: 0000000000000001
[ 22.757508] RAX: ffffffffffffffda RBX: 0000000000000002 RCX: 00007f14a697a240
[ 22.757813] RDX: 0000000000000002 RSI: 0000558b25aeda50 RDI: 0000000000000001
[ 22.758117] RBP: 0000558b25aeda50 R08: 00007f14a6a54d90 R09: 00007f14a6a54d90
[ 22.758422] R10: 0000000000000000 R11: 0000000000000202 R12: 0000000000000002
[ 22.758726] R13: 00007f14a6a55760 R14: 0000000000000002 R15: 00007f14a6a509e0
[ 22.759039] </TASK>
> return;
>
> device_del(&dev->dev);
>--
>2.43.5
>
next prev parent reply other threads:[~2024-10-03 2:34 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-08-27 19:28 [PATCHv2 0/5] pci cleanup/prep patches Keith Busch
2024-08-27 19:28 ` [PATCHv2 1/5] pci: make pci_stop_dev concurrent safe Keith Busch
2024-10-02 23:39 ` Davidlohr Bueso
2024-10-03 0:04 ` Keith Busch
2024-08-27 19:28 ` [PATCHv2 2/5] pci: make pci_destroy_dev " Keith Busch
2024-10-03 2:34 ` Davidlohr Bueso [this message]
2024-10-03 14:54 ` Keith Busch
2024-10-03 17:04 ` Davidlohr Bueso
2024-10-03 17:59 ` Keith Busch
2024-10-08 2:15 ` Davidlohr Bueso
2024-10-22 20:29 ` Keith Busch
2024-08-27 19:28 ` [PATCHv2 3/5] pci: move the walk bus lock to where its needed Keith Busch
2024-10-03 0:32 ` Davidlohr Bueso
2024-10-09 11:09 ` Ilpo Järvinen
2024-08-27 19:28 ` [PATCHv2 4/5] pci: walk bus recursively Keith Busch
2024-10-09 12:08 ` Ilpo Järvinen
2024-08-27 19:28 ` [PATCHv2 5/5] pci: unexport pci_walk_bus_locked Keith Busch
2024-10-03 0:35 ` Davidlohr Bueso
2024-10-09 12:20 ` Ilpo Järvinen
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20241003023354.txfw7w4ud247h5va@offworld \
--to=dave@stgolabs.net \
--cc=Jonathan.Cameron@huawei.com \
--cc=bhelgaas@google.com \
--cc=kbusch@kernel.org \
--cc=kbusch@meta.com \
--cc=linux-pci@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox