[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <fa90e815-daff-57b1-7930-447c6f5fa3ab@users.sourceforge.net>
Date: Wed, 5 Apr 2017 16:54:40 +0200
From: SF Markus Elfring <elfring@...rs.sourceforge.net>
To: Yuval Shaia <yuval.shaia@...cle.com>
Cc: linux-rdma@...r.kernel.org, Doug Ledford <dledford@...hat.com>,
Hal Rosenstock <hal.rosenstock@...il.com>,
Mike Marciniszyn <infinipath@...el.com>,
Sean Hefty <sean.hefty@...el.com>,
LKML <linux-kernel@...r.kernel.org>,
kernel-janitors@...r.kernel.org
Subject: Re: [PATCH 1/5] IB/qib: Use kcalloc() in qib_init_iba7322_funcs()
>> @@ -7324,8 +7324,9 @@ struct qib_devdata *qib_init_iba7322_funcs(struct pci_dev *pdev,
>> actual_cnt -= dd->num_pports;
>>
>> tabsize = actual_cnt;
>> - dd->cspec->msix_entries = kzalloc(tabsize *
>> - sizeof(struct qib_msix_entry), GFP_KERNEL);
>> + dd->cspec->msix_entries = kcalloc(tabsize,
>> + sizeof(*dd->cspec->msix_entries),
>> + GFP_KERNEL);
>
> Are we fine with loosing the zeroing of the entries?
How did you get this concern?
Do you really miss such functionality from the other interface?
Regards,
Markus
Powered by blists - more mailing lists