From: Mahesh Rao <mahesh.rao@altera.com>
To: Greg KH <gregkh@linuxfoundation.org>
Cc: Dinh Nguyen <dinguyen@kernel.org>,
linux-kernel@vger.kernel.org,
Matthew Gerlach <matthew.gerlach@altera.com>
Subject: Re: [PATCH 1/5] firmware: stratix10-svc: Add mutex lock and unlock in stratix10 memory allocation/free
Date: Tue, 26 Aug 2025 09:19:29 +0530 [thread overview]
Message-ID: <7276113b-28bb-4347-9e7d-6b63fb03a36b@altera.com> (raw)
In-Reply-To: <2025082234-scarcity-relive-9362@gregkh>
On 22-08-2025 05:19 pm, Greg KH wrote:
> On Fri, Aug 22, 2025 at 03:17:54PM +0530, Mahesh Rao wrote:
>> Hi Greg,
>> thanks for reviewing the code.
>>
>> On 19-08-2025 04:36 pm, Greg KH wrote:
>>> On Tue, Jul 22, 2025 at 11:30:41AM -0500, Dinh Nguyen wrote:
>>>> From: Mahesh Rao <mahesh.rao@altera.com>
>>>>
>>>> This commit adds a mutex lock to protect the
>>>> stratix10_svc_allocate_memory and
>>>> stratix10_svc_free_memory functions to ensure
>>>> thread safety when allocating and freeing memory.
>>>> This prevents potential race conditions and ensures
>>>> synchronization.
>>>
>>> You have 72 columns to write a changelog in, please use it :)
>>>
>>> And is this fixing a bug? If so, shouldn't this be tagged for stable
>>> and add a Fixes: tag?
>>>
>>> If this isn't a bug, then why is it needed? How can these race?
>>
>> In the current implementation, all operations were performed serially,
>> eliminating the need for protection mechanisms. However, with this patch
>> set, we are introducing parallel access and communication with the SDM
>> across multiple client drivers. This change may lead to race conditions
>> involving the svc_data_mem list.
>
> Then that needs to be said here :)
>
> Also, what is causing these operations to be performed serially if there
> is no locking?
Use case wise, currently we are only accessing this serially. But I
agree that there is a scenario where parallel access is possible. So I
shall add fixes tag in the next revision.
>
>>
>>>
>>>>
>>>> Signed-off-by: Mahesh Rao <mahesh.rao@altera.com>
>>>> Reviewed-by: Matthew Gerlach <matthew.gerlach@altera.com>
>>>> Signed-off-by: Dinh Nguyen <dinguyen@kernel.org>
>>>> ---
>>>> drivers/firmware/stratix10-svc.c | 31 ++++++++++++++++++++++++-------
>>>> 1 file changed, 24 insertions(+), 7 deletions(-)
>>>>
>>>> diff --git a/drivers/firmware/stratix10-svc.c b/drivers/firmware/stratix10-svc.c
>>>> index e3f990d888d7..73c77b8e9f2b 100644
>>>> --- a/drivers/firmware/stratix10-svc.c
>>>> +++ b/drivers/firmware/stratix10-svc.c
>>>> @@ -1,6 +1,7 @@
>>>> // SPDX-License-Identifier: GPL-2.0
>>>> /*
>>>> * Copyright (C) 2017-2018, Intel Corporation
>>>> + * Copyright (C) 2025, Altera Corporation
>>>> */
>>>> #include <linux/completion.h>
>>>> @@ -171,6 +172,10 @@ struct stratix10_svc_chan {
>>>> static LIST_HEAD(svc_ctrl);
>>>> static LIST_HEAD(svc_data_mem);
>>>> +/* svc_mem_lock protects access to the svc_data_mem list for
>>>> + * concurrent multi-client operations
>>>> + */
>>>
>>> Odd coding style, this isn't the network subsystem :(
>>
>> Ok sure, will change
>>
>>>
>>> And what about a lock for svc_ctrl?
>>
>> There is only one instance of svc_ctrl and there is no parallel access to
>> it.so a lock is not required as of now.
>
> But don't you have multiple places that list can be accessed now at the
> same time?
>
> In other words, what is changing to require one list to require it but
> not the other? Is there some other lock for that?
This is just a context controller having only 1 instance as of now and
this context object is only updated during driver init and read in
another function.
>
> thanks,
>
> greg k-h
Thanks
Mahesh
prev parent reply other threads:[~2025-08-26 3:49 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-07-22 16:30 [PATCH 1/5] firmware: stratix10-svc: Add mutex lock and unlock in stratix10 memory allocation/free Dinh Nguyen
2025-07-22 16:30 ` [PATCH 2/5] firmware: stratix10-svc: Implement ID pool management for asynchronous operations Dinh Nguyen
2025-08-19 11:07 ` Greg KH
2025-08-22 9:48 ` Mahesh Rao
2025-07-22 16:30 ` [PATCH 3/5] firmware: stratix10-svc: Add initial support for asynchronous communication with Stratix10 service channel Dinh Nguyen
2025-08-19 11:08 ` Greg KH
2025-08-22 9:48 ` Mahesh Rao
2025-07-22 16:30 ` [PATCH 4/5] firmware: stratix10-svc: Add support for RSU commands in asynchronous framework Dinh Nguyen
2025-07-22 16:30 ` [PATCH 5/5] firmware: stratix10-rsu: Migrate RSU driver to use stratix10 " Dinh Nguyen
2025-08-19 11:06 ` [PATCH 1/5] firmware: stratix10-svc: Add mutex lock and unlock in stratix10 memory allocation/free Greg KH
2025-08-22 9:47 ` Mahesh Rao
2025-08-22 11:49 ` Greg KH
2025-08-26 3:49 ` Mahesh Rao [this message]
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=7276113b-28bb-4347-9e7d-6b63fb03a36b@altera.com \
--to=mahesh.rao@altera.com \
--cc=dinguyen@kernel.org \
--cc=gregkh@linuxfoundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=matthew.gerlach@altera.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;
as well as URLs for NNTP newsgroup(s).