mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Haakon Bugge <haakon.bugge@oracle.com>
To: Greg KH <gregkh@linuxfoundation.org>
Cc: Jason Gunthorpe <jgg@nvidia.com>,
	Leon Romanovsky <leon@kernel.org>,
	Doug Ledford <dledford@redhat.com>,
	Kees Cook <keescook@chromium.org>,
	Nathan Chancellor <nathan@kernel.org>,
	Adit Ranadive <aditr@vmware.com>,
	Ariel Elior <aelior@marvell.com>,
	Christian Benvenuti <benve@cisco.com>,
	"clang-built-linux@googlegroups.com" 
	<clang-built-linux@googlegroups.com>,
	Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>,
	Devesh Sharma <devesh.sharma@broadcom.com>,
	Gal Pressman <galpress@amazon.com>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	OFED mailing list <linux-rdma@vger.kernel.org>,
	Michal Kalderon <mkalderon@marvell.com>,
	Mike Marciniszyn <mike.marciniszyn@cornelisnetworks.com>,
	Mustafa Ismail <mustafa.ismail@intel.com>,
	Naresh Kumar PBS <nareshkumar.pbs@broadcom.com>,
	Nelson Escobar <neescoba@cisco.com>,
	Nick Desaulniers <ndesaulniers@google.com>,
	Potnuri Bharat Teja <bharat@chelsio.com>,
	Selvin Xavier <selvin.xavier@broadcom.com>,
	Shiraz Saleem <shiraz.saleem@intel.com>,
	VMware PV-Drivers <pv-drivers@vmware.com>,
	Yishai Hadas <yishaih@nvidia.com>,
	Zhu Yanjun <zyjzyj2000@gmail.com>
Subject: Re: [PATCH rdma-next v1 10/15] RDMA/cm: Use an attribute_group on the ib_port_attribute intead of kobj's
Date: Mon, 14 Jun 2021 03:27:00 +0000	[thread overview]
Message-ID: <92A96585-DAE5-46C5-8D2A-2EED92F75FDF@oracle.com> (raw)
In-Reply-To: <YMMb9NZ0nHRTullc@kroah.com>



> On 11 Jun 2021, at 10:16, Greg KH <gregkh@linuxfoundation.org> wrote:
> 
> On Fri, Jun 11, 2021 at 07:25:46AM +0000, Haakon Bugge wrote:
>> 
>> 
>>> On 7 Jun 2021, at 14:50, Jason Gunthorpe <jgg@nvidia.com> wrote:
>>> 
>>> On Mon, Jun 07, 2021 at 02:39:45PM +0200, Greg KH wrote:
>>>> On Mon, Jun 07, 2021 at 09:14:11AM -0300, Jason Gunthorpe wrote:
>>>>> On Mon, Jun 07, 2021 at 12:25:03PM +0200, Greg KH wrote:
>>>>>> On Mon, Jun 07, 2021 at 11:17:35AM +0300, Leon Romanovsky wrote:
>>>>>>> From: Jason Gunthorpe <jgg@nvidia.com>
>>>>>>> 
>>>>>>> This code is trying to attach a list of counters grouped into 4 groups to
>>>>>>> the ib_port sysfs. Instead of creating a bunch of kobjects simply express
>>>>>>> everything naturally as an ib_port_attribute and add a single
>>>>>>> attribute_groups list.
>>>>>>> 
>>>>>>> Remove all the naked kobject manipulations.
>>>>>> 
>>>>>> Much nicer.
>>>>>> 
>>>>>> But why do you need your counters to be atomic in the first place?  What
>>>>>> are they counting that requires this?  
>>>>> 
>>>>> The write side of the counter is being updated from concurrent kernel
>>>>> threads without locking, so this is an atomic because the write side
>>>>> needs atomic_add().
>>>> 
>>>> So the atomic write forces a lock :(
>>> 
>>> Of course, but a single atomic is cheaper than the double atomic in a
>>> full spinlock.
>>> 
>>>>> Making them a naked u64 will cause significant corruption on the write
>>>>> side, and packet counters that are not accurate after quiescence are
>>>>> not very useful things.
>>>> 
>>>> How "accurate" do these have to be?
>>> 
>>> They have to be accurate. They are networking packet counters. What is
>>> the point of burning CPU cycles keeping track of inaccurate data?
>> 
>> Consider a CPU with a 32-bit wide datapath to memory, which reads and writes the most significant 4-byte word first:
> 
> What CPU is that?

Hypothetical32 :-)

>>    Memory                   CPU1                   CPU2
>> MSW         LSW        MSW         LSW        MSW         LSW
>> 0x0  0xffffffff
>> 0x0  0xffffffff        0x0
>> 0x0  0xffffffff        0x0  0xffffffff
>> 0x0  0xffffffff        0x1         0x0                         cpu1 has incremented its register
>> 0x1  0xffffffff        0x1         0x0                         cpu1 has written msw
>> 0x1  0xffffffff        0x1         0x0        0x1              cpu2 has read msw
>> 0x1  0xffffffff        0x1         0x0        0x1  0xffffffff
>> 0x1         0x0        0x1         0x0        0x2         0x0
>> 0x2         0x0        0x1         0x0        0x2         0x0
>> 0x2         0x0        0x1         0x0        0x2         0x0
>> 
>> 
>> I would say that 0x200000000 vs. 0x100000001 is more than inaccurate!
> 
> True, then maybe these should just be 32bit counters :)

How long can we then run without wrapping? Our UEK is security updated by means of ksplice, and since the introduction of Spectre/Meltdown CPU fixes in 2018, we have been able to update the kernels running wrt. security fixes.

I see no harm by using an atomic 64-bit add. Yes, it serializes the pipeline and locks the cache-line in the first-level cache for as long as it takes to for a RMW on it. Compared to surround an ordinary add with lock/unlock, an atomic increment is strongly preferred in my opinion.

Ordinary add without locking leads to the issue above on systems with 32-bit wide memory data paths.

Using 32-bit counters raises the issue of wrapping in systems running for years and having a high frequency of IB connection forming and resurrections.


Thxs, Håkon


> thanks,
> 
> greg k-h


  reply	other threads:[~2021-06-14  3:32 UTC|newest]

Thread overview: 38+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-06-07  8:17 [PATCH rdma-next v1 00/15] Reorganize sysfs file creation for struct ib_devices Leon Romanovsky
2021-06-07  8:17 ` [PATCH rdma-next v1 01/15] RDMA: Split the alloc_hw_stats() ops to port and device variants Leon Romanovsky
2021-06-07  8:17 ` [PATCH rdma-next v1 02/15] RDMA/core: Replace the ib_port_data hw_stats pointers with a ib_port pointer Leon Romanovsky
2021-06-07 10:23   ` Greg KH
2021-06-07 11:00     ` Leon Romanovsky
2021-06-07 11:23       ` Greg KH
2021-06-07 12:32     ` Jason Gunthorpe
2021-06-07  8:17 ` [PATCH rdma-next v1 03/15] RDMA/core: Split port and device counter sysfs attributes Leon Romanovsky
2021-06-07  8:17 ` [PATCH rdma-next v1 04/15] RDMA/core: Split gid_attrs related sysfs from add_port() Leon Romanovsky
2021-06-07  8:17 ` [PATCH rdma-next v1 05/15] RDMA/core: Simplify how the gid_attrs sysfs is created Leon Romanovsky
2021-06-07  8:17 ` [PATCH rdma-next v1 06/15] RDMA/core: Simplify how the port " Leon Romanovsky
2021-06-07  8:17 ` [PATCH rdma-next v1 07/15] RDMA/core: Create the device hw_counters through the normal groups mechanism Leon Romanovsky
2021-06-07  8:17 ` [PATCH rdma-next v1 08/15] RDMA/core: Remove the kobject_uevent() NOP Leon Romanovsky
2021-06-07  8:17 ` [PATCH rdma-next v1 09/15] RDMA/core: Expose the ib port sysfs attribute machinery Leon Romanovsky
2021-06-07  8:17 ` [PATCH rdma-next v1 10/15] RDMA/cm: Use an attribute_group on the ib_port_attribute intead of kobj's Leon Romanovsky
2021-06-07 10:25   ` Greg KH
2021-06-07 10:51     ` Leon Romanovsky
2021-06-07 11:22       ` Greg KH
2021-06-07 11:37         ` Leon Romanovsky
2021-06-07 12:08           ` Greg KH
2021-06-07 12:41             ` Leon Romanovsky
2021-06-07 12:14     ` Jason Gunthorpe
2021-06-07 12:39       ` Greg KH
2021-06-07 12:50         ` Jason Gunthorpe
2021-06-11  7:25           ` Haakon Bugge
2021-06-11  8:16             ` Greg KH
2021-06-14  3:27               ` Haakon Bugge [this message]
2021-06-07  8:17 ` [PATCH rdma-next v1 11/15] RDMA/qib: Use attributes for the port sysfs Leon Romanovsky
2021-06-07  8:17 ` [PATCH rdma-next v1 12/15] RDMA/hfi1: " Leon Romanovsky
2021-06-07  8:17 ` [PATCH rdma-next v1 13/15] RDMA: Change ops->init_port to ops->port_groups Leon Romanovsky
2021-06-07  8:17 ` [PATCH rdma-next v1 14/15] RDMA/core: Allow port_groups to be used with namespaces Leon Romanovsky
2021-06-07 13:29   ` Parav Pandit
2021-06-07 23:32     ` Jason Gunthorpe
2021-06-08  5:29       ` Parav Pandit
2021-06-08  4:55     ` Leon Romanovsky
2021-06-08  5:32       ` Parav Pandit
2021-06-08  6:57         ` Leon Romanovsky
2021-06-07  8:17 ` [PATCH rdma-next v1 15/15] RDMA: Remove rdma_set_device_sysfs_group() Leon Romanovsky

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=92A96585-DAE5-46C5-8D2A-2EED92F75FDF@oracle.com \
    --to=haakon.bugge@oracle.com \
    --cc=aditr@vmware.com \
    --cc=aelior@marvell.com \
    --cc=benve@cisco.com \
    --cc=bharat@chelsio.com \
    --cc=clang-built-linux@googlegroups.com \
    --cc=dennis.dalessandro@cornelisnetworks.com \
    --cc=devesh.sharma@broadcom.com \
    --cc=dledford@redhat.com \
    --cc=galpress@amazon.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=jgg@nvidia.com \
    --cc=keescook@chromium.org \
    --cc=leon@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-rdma@vger.kernel.org \
    --cc=mike.marciniszyn@cornelisnetworks.com \
    --cc=mkalderon@marvell.com \
    --cc=mustafa.ismail@intel.com \
    --cc=nareshkumar.pbs@broadcom.com \
    --cc=nathan@kernel.org \
    --cc=ndesaulniers@google.com \
    --cc=neescoba@cisco.com \
    --cc=pv-drivers@vmware.com \
    --cc=selvin.xavier@broadcom.com \
    --cc=shiraz.saleem@intel.com \
    --cc=yishaih@nvidia.com \
    --cc=zyjzyj2000@gmail.com \
    /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

all inboxes | Powered by JetHome®