[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20250314082527.GU5880@noisy.programming.kicks-ass.net>
Date: Fri, 14 Mar 2025 09:25:27 +0100
From: Peter Zijlstra <peterz@...radead.org>
To: Shrikanth Hegde <sshegde@...ux.ibm.com>
Cc: maddy@...ux.ibm.com, linuxppc-dev@...ts.ozlabs.org, npiggin@...il.com,
christophe.leroy@...roup.eu, mpe@...erman.id.au,
fbarrat@...ux.ibm.com, ajd@...ux.ibm.com, mahesh@...ux.ibm.com,
oohall@...il.com, hbathini@...ux.ibm.com, dhowells@...hat.com,
haren@...ux.ibm.com, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 4/6] powerpc: book3s: vas: use lock guard for mutex
On Fri, Mar 14, 2025 at 11:15:42AM +0530, Shrikanth Hegde wrote:
> use guard(mutex) for scope based resource management of mutex.
> This would make the code simpler and easier to maintain.
>
> More details on lock guards can be found at
> https://lore.kernel.org/all/20230612093537.614161713@infradead.org/T/#u
>
> There is also an example of using scoped_guard.
>
> Signed-off-by: Shrikanth Hegde <sshegde@...ux.ibm.com>
> ---
> arch/powerpc/platforms/book3s/vas-api.c | 19 ++++++-------------
> 1 file changed, 6 insertions(+), 13 deletions(-)
>
> diff --git a/arch/powerpc/platforms/book3s/vas-api.c b/arch/powerpc/platforms/book3s/vas-api.c
> index 0b6365d85d11..eb1a97271afb 100644
> --- a/arch/powerpc/platforms/book3s/vas-api.c
> +++ b/arch/powerpc/platforms/book3s/vas-api.c
> @@ -425,7 +425,7 @@ static vm_fault_t vas_mmap_fault(struct vm_fault *vmf)
> return VM_FAULT_SIGBUS;
> }
>
> - mutex_lock(&txwin->task_ref.mmap_mutex);
> + guard(mutex)(&txwin->task_ref.mmap_mutex);
> /*
> * The window may be inactive due to lost credit (Ex: core
> * removal with DLPAR). If the window is active again when
> @@ -437,11 +437,9 @@ static vm_fault_t vas_mmap_fault(struct vm_fault *vmf)
> if (paste_addr) {
> fault = vmf_insert_pfn(vma, vma->vm_start,
> (paste_addr >> PAGE_SHIFT));
> - mutex_unlock(&txwin->task_ref.mmap_mutex);
> return fault;
> }
> }
> - mutex_unlock(&txwin->task_ref.mmap_mutex);
I had to open up this file to check, but this seems incorrect since you
now also run do_fail_paste() with the lock held, where previously you
did not.
> /*
> * Received this fault due to closing the actual window.
Powered by blists - more mailing lists