linux-kernel - Re: [PATCH] staging: r8188eu: Remove _enter/_exit_critical

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 for Android: free password hash cracker in your pocket

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

Message-ID: <2072985.qfodHclbUl@localhost.localdomain>
Date:   Thu, 19 Aug 2021 12:16:03 +0200
From:   "Fabio M. De Francesco" <fmdefrancesco@...il.com>
To:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Cc:     Larry Finger <Larry.Finger@...inger.net>,
        Phillip Potter <phil@...lpotter.co.uk>,
        Martin Kaiser <martin@...ser.cx>,
        Michael Straube <straube.linux@...il.com>,
        linux-staging@...ts.linux.dev, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] staging: r8188eu: Remove _enter/_exit_critical_mutex()

On Thursday, August 19, 2021 9:07:20 AM CEST Fabio M. De Francesco wrote:
> On Thursday, August 19, 2021 8:30:21 AM CEST Greg Kroah-Hartman wrote:
> > On Thu, Aug 19, 2021 at 08:08:37AM +0200, Fabio M. De Francesco wrote:
> > > Remove _enter_critical_mutex() and _exit_critical_mutex(). They are
> > > unnecessary wrappers, respectively to mutex_lock_interruptible and to
> > > mutex_unlock(). They also have an odd interface that takes an unused
> > > second parameter "unsigned long *pirqL".
> > > 
> > > Use directly the in-kernel API; check and manage the return value of
> > > mutex_lock_interruptible().
> > > 
> > > Signed-off-by: Fabio M. De Francesco <fmdefrancesco@...il.com>
> > > ---
> > >  drivers/staging/r8188eu/core/rtw_mlme_ext.c     |  5 +++--
> > >  drivers/staging/r8188eu/hal/usb_ops_linux.c     |  7 +++++--
> > >  drivers/staging/r8188eu/include/osdep_service.h | 13 -------------
> > >  drivers/staging/r8188eu/os_dep/os_intfs.c       |  5 +++--
> > >  4 files changed, 11 insertions(+), 19 deletions(-)
> > > 
> > > diff --git a/drivers/staging/r8188eu/core/rtw_mlme_ext.c b/drivers/staging/r8188eu/core/rtw_mlme_ext.c
> > > index f6ee72d5af09..484083468ebb 100644
> > > --- a/drivers/staging/r8188eu/core/rtw_mlme_ext.c
> > > +++ b/drivers/staging/r8188eu/core/rtw_mlme_ext.c
> > > @@ -4358,7 +4358,8 @@ s32 dump_mgntframe_and_wait_ack(struct adapter *padapter, struct xmit_frame *pmg
> > >  	if (padapter->bSurpriseRemoved || padapter->bDriverStopped)
> > >  		return -1;
> > >  
> > > -	_enter_critical_mutex(&pxmitpriv->ack_tx_mutex, NULL);
> > > +	if (mutex_lock_interruptible(&pxmitpriv->ack_tx_mutex))
> > > +		return -EINTR;
> > 
> > But the code never would return this value if the lock function returned
> > an error.  Why do that here now?

Ah, now I think I understand what you asked me ... sorry for not having 
immediately grasped the meaning of your objection. :(

I guess you wanted to know why I decided to check and handle the 
return values of mutex_lock_interruptible (), as the original code didn't. 
Did I understand the correct meaning of your question?

If so, now I can explain why I did it ...

A few months ago I did the conversion of the visorhba (Unisys) driver from 
IDR to XArray. Since the old code did not check IDR API return values, I had 
decided not to check for XArray API return values as well.

Then Dan C. asked me to implement the checks that were missing in the 
original code. So, today I decided to implement them before I was asked 
to do it. Now it's clear that in this case they are not needed.

That's all. :-)

I'm about to send a v2 without those unnecessary checks.

Thanks,

Fabio

> I read from the documentation that "[mutex_lock_interruptible()] Return: 0 if 
> the lock was successfully acquired or -EINTR if a signal arrived.". 
> 
> After reading that, I thought that if I got -EINTR I should return it. Shouldn't I?
> 
> Now I've just checked its usage pattern in another file where we have exactly 12
> times the same management of the error (the example I'm talking about is in
> drivers/staging/vc04_services/vchiq-mmal/mmal-vchiq.c):
> 
> "if (mutex_lock_interruptible(&instance->vchiq_mutex))
>                 return -EINTR;".
> 
> Unless you mean that I should return the "ret" variable, which is already set to 
> "_FAIL", I am really confused. Please, can you further elaborate what I'm doing 
> wrong?
> 
> Thanks,
> 
> Fabio