Ping Herbert. Thanks. > -----Original Message----- > From: Gonglei (Arei) > Sent: Wednesday, September 27, 2023 5:18 PM > To: 'Halil Pasic' <pasic@xxxxxxxxxxxxx> > Cc: Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx>; linux-crypto@xxxxxxxxxxxxxxx; > Marc Hartmayer <mhartmay@xxxxxxxxxxxxx>; Michael S. Tsirkin > <mst@xxxxxxxxxx>; Jason Wang <jasowang@xxxxxxxxxx>; > virtualization@xxxxxxxxxxxxxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx; > pizhenwei@xxxxxxxxxxxxx; Cornelia Huck <cohuck@xxxxxxxxxx> > Subject: RE: [PATCH] crypto: virtio-crypto: call finalize with bh disabled > > > > > -----Original Message----- > > From: Halil Pasic [mailto:pasic@xxxxxxxxxxxxx] > > Sent: Wednesday, September 27, 2023 12:42 AM > > To: Gonglei (Arei) <arei.gonglei@xxxxxxxxxx> > > Cc: Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx>; > > linux-crypto@xxxxxxxxxxxxxxx; Marc Hartmayer <mhartmay@xxxxxxxxxxxxx>; > > Michael S. Tsirkin <mst@xxxxxxxxxx>; Jason Wang > <jasowang@xxxxxxxxxx>; > > virtualization@xxxxxxxxxxxxxxxxxxxxxxxxxx; > > linux-kernel@xxxxxxxxxxxxxxx; pizhenwei@xxxxxxxxxxxxx; Halil Pasic > > <pasic@xxxxxxxxxxxxx>; Cornelia Huck <cohuck@xxxxxxxxxx> > > Subject: Re: [PATCH] crypto: virtio-crypto: call finalize with bh > > disabled > > > > [..] > > > --- a/drivers/crypto/virtio/virtio_crypto_akcipher_algs.c > > > +++ b/drivers/crypto/virtio/virtio_crypto_akcipher_algs.c > > > @@ -61,8 +61,9 @@ static void virtio_crypto_akcipher_finalize_req( > > > vc_akcipher_req->src_buf = NULL; > > > vc_akcipher_req->dst_buf = NULL; > > > virtcrypto_clear_request(&vc_akcipher_req->base); > > > - > > > + local_bh_disable(); > > > > > > crypto_finalize_akcipher_request(vc_akcipher_req->base.dataq->engine > > > , > > > req, err); > > > + local_bh_enable(); > > > > Thanks Gonglei! > > > > I did this a quick spin, and it does not seem to be sufficient on s390x. > > Which does not come as a surprise to me, because > > > > #define lockdep_assert_in_softirq() > > \ > > do > > { > > \ > > WARN_ON_ONCE(__lockdep_enabled && > > \ > > (!in_softirq() || in_irq() || in_nmi())); \ > > } while (0) > > > > will still warn because in_irq() still evaluates to true (your patch > > addresses the !in_softirq() part). > > > You are right. > > So I think the core of this question is: Can we call crypto_finalize_request() in > the upper half of the interrupt? > If so, maybe we should introduce a new function, such as > lockdep_assert_in_interrupt(). > > #define lockdep_assert_in_interrupt() \ > do { \ > WARN_ON_ONCE(__lockdep_enabled && !in_interrupt()); \ > } while (0) > > If not, why? > > Herbert, do you have any suggestions? Thanks. > > > Regards, > -Gonglei >
