linux-kernel - Re: [PATCH] pci: call _cond_resched() after pci_bus_write

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20211013190014.GA1909934@bhelgaas>
Date:   Wed, 13 Oct 2021 14:00:14 -0500
From:   Bjorn Helgaas <helgaas@...nel.org>
To:     menglong8.dong@...il.com
Cc:     bhelgaas@...gle.com, linux-pci@...r.kernel.org,
        linux-kernel@...r.kernel.org, Menglong Dong <imagedong@...cent.com>
Subject: Re: [PATCH] pci: call _cond_resched() after pci_bus_write_config

Match previous subject lines (use "git log --oneline
drivers/pci/access.c" to see them).

On Wed, Oct 13, 2021 at 08:55:42PM +0800, menglong8.dong@...il.com wrote:
> From: Menglong Dong <imagedong@...cent.com>
> 
> While the system is running in KVM, pci config writing for virtio devices
> may cost long time(about 1-2ms), as it causes VM-exit. During
> __pci_bus_assign_resources(), pci_setup_bridge, which can do pci config
> writing up to 10 times, can be called many times without any
> _cond_resched(). So __pci_bus_assign_resources can cause 25+ms scheduling
> latency with !CONFIG_PREEMPT.
> 
> To solve this problem, call _cond_resched() after pci config writing.

s/pci/PCI/ above.
Add space before "(".
Add "()" after function names consistently (some have it, some don't).

What exactly is the problem?  I expect __pci_bus_assign_resources() to
be used mostly during boot-time enumeration.  How much of a problem is
the latency at that point?  Why is this particularly a problem in the
KVM environment?  Or is it also a problem on bare metal?

Are there other config write paths that should have a similar change?

_cond_resched() only appears here:

  $ git grep "\<_cond_resched\>"
  include/linux/sched.h:static __always_inline int _cond_resched(void)
  include/linux/sched.h:static inline int _cond_resched(void)
  include/linux/sched.h:static inline int _cond_resched(void) { return 0; }
  include/linux/sched.h:  _cond_resched();

so I don't believe PCI is so special that this needs to be the only
other use.  Maybe a different resched interface is more appropriate?

> Signed-off-by: Menglong Dong <imagedong@...cent.com>
> ---
>  drivers/pci/access.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/pci/access.c b/drivers/pci/access.c
> index 46935695cfb9..babed43702df 100644
> --- a/drivers/pci/access.c
> +++ b/drivers/pci/access.c
> @@ -57,6 +57,7 @@ int noinline pci_bus_write_config_##size \
>  	pci_lock_config(flags);						\
>  	res = bus->ops->write(bus, devfn, pos, len, value);		\
>  	pci_unlock_config(flags);					\
> +	_cond_resched();						\
>  	return res;							\
>  }
>  
> -- 
> 2.27.0
>