From mboxrd@z Thu Jan 1 00:00:00 1970 From: Pavel Begunkov Subject: Re: IORING_REGISTER_CREDS[_UPDATE]() and credfd_create()? Date: Wed, 29 Jan 2020 03:10:00 +0300 Message-ID: References: <688e187a-75dd-89d9-921c-67de228605ce@samba.org> <1ac31828-e915-6180-cdb4-36685442ea75@kernel.dk> <0d4f43d8-a0c4-920b-5b8f-127c1c5a3fad@kernel.dk> <2d7e7fa2-e725-8beb-90b9-6476d48bdb33@gmail.com> <6c401e23-de7c-1fc1-4122-33d53fcf9700@kernel.dk> <35eebae7-76dd-52ee-58b2-4f9e85caee40@kernel.dk> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha256; protocol="application/pgp-signature"; boundary="lRrc6G6gXrup0csTFFZcEntNplhztBpnq" Return-path: In-Reply-To: <35eebae7-76dd-52ee-58b2-4f9e85caee40-tSWWG44O7X1aa/9Udqfwiw@public.gmane.org> Sender: linux-api-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Jens Axboe , Stefan Metzmacher Cc: io-uring , Linux API Mailing List List-Id: linux-api@vger.kernel.org This is an OpenPGP/MIME signed message (RFC 4880 and 3156) --lRrc6G6gXrup0csTFFZcEntNplhztBpnq Content-Type: multipart/mixed; boundary="cK8FbgRep79RadR9iuAsWyEL8b1jMSHXS"; protected-headers="v1" From: Pavel Begunkov To: Jens Axboe , Stefan Metzmacher Cc: io-uring , Linux API Mailing List Message-ID: Subject: Re: IORING_REGISTER_CREDS[_UPDATE]() and credfd_create()? References: <688e187a-75dd-89d9-921c-67de228605ce-eUNUBHrolfbYtjvyW6yDsg@public.gmane.org> <1ac31828-e915-6180-cdb4-36685442ea75-tSWWG44O7X1aa/9Udqfwiw@public.gmane.org> <0d4f43d8-a0c4-920b-5b8f-127c1c5a3fad-tSWWG44O7X1aa/9Udqfwiw@public.gmane.org> <2d7e7fa2-e725-8beb-90b9-6476d48bdb33-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> <6c401e23-de7c-1fc1-4122-33d53fcf9700-tSWWG44O7X1aa/9Udqfwiw@public.gmane.org> <35eebae7-76dd-52ee-58b2-4f9e85caee40-tSWWG44O7X1aa/9Udqfwiw@public.gmane.org> In-Reply-To: <35eebae7-76dd-52ee-58b2-4f9e85caee40-tSWWG44O7X1aa/9Udqfwiw@public.gmane.org> --cK8FbgRep79RadR9iuAsWyEL8b1jMSHXS Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: quoted-printable On 29/01/2020 02:51, Jens Axboe wrote: > On 1/28/20 4:40 PM, Jens Axboe wrote: >> On 1/28/20 4:36 PM, Pavel Begunkov wrote: >>> On 28/01/2020 22:42, Jens Axboe wrote: >>>> I didn't like it becoming a bit too complicated, both in terms of >>>> implementation and use. And the fact that we'd have to jump through >>>> hoops to make this work for a full chain. >>>> >>>> So I punted and just added sqe->personality and IOSQE_PERSONALITY. >>>> This makes it way easier to use. Same branch: >>>> >>>> https://git.kernel.dk/cgit/linux-block/log/?h=3Dfor-5.6/io_uring-vfs= -creds >>>> >>>> I'd feel much better with this variant for 5.6. >>>> >>> >>> Checked out ("don't use static creds/mm assignments") >>> >>> 1. do we miscount cred refs? We grab one in get_current_cred() for ea= ch async >>> request, but if (worker->creds !=3D work->creds) it will never be put= =2E >> >> Yeah I think you're right, that needs a bit of fixing up. >=20 Hmm, it seems it leaks it unconditionally, as it grabs in a ref in overri= de_creds(). > I think this may have gotten fixed with the later addition posted today= ? > I'll double check. But for the newer stuff, we put it for both cases > when the request is freed. Yeah, maybe. I got tangled trying to verify both at once and decided to s= tart with the old one. >>> 2. shouldn't worker->creds be named {old,saved,etc}_creds? It's set a= s >>> >>> worker->creds =3D override_creds(work->creds); >>> >>> Where override_creds() returns previous creds. And if so, then the fo= llowing >>> fast check looks strange: >>> >>> worker->creds !=3D work->creds >> >> Don't care too much about the naming, but the logic does appear off. >> I'll take a look at both of these tonight, unless you beat me to it. Apparently, you're faster :) >=20 > Testing this now, what a braino. >=20 > diff --git a/fs/io-wq.c b/fs/io-wq.c > index ee49e8852d39..8fbbadf04cc3 100644 > --- a/fs/io-wq.c > +++ b/fs/io-wq.c > @@ -56,7 +56,8 @@ struct io_worker { > =20 > struct rcu_head rcu; > struct mm_struct *mm; > - const struct cred *creds; > + const struct cred *cur_creds; > + const struct cred *saved_creds; > struct files_struct *restore_files; > }; > =20 > @@ -135,9 +136,9 @@ static bool __io_worker_unuse(struct io_wqe *wqe, s= truct io_worker *worker) > { > bool dropped_lock =3D false; > =20 > - if (worker->creds) { > - revert_creds(worker->creds); > - worker->creds =3D NULL; > + if (worker->saved_creds) { > + revert_creds(worker->saved_creds); > + worker->cur_creds =3D worker->saved_creds =3D NULL; > } > =20 > if (current->files !=3D worker->restore_files) { > @@ -424,10 +425,11 @@ static void io_wq_switch_mm(struct io_worker *wor= ker, struct io_wq_work *work) > static void io_wq_switch_creds(struct io_worker *worker, > struct io_wq_work *work) > { > - if (worker->creds) > - revert_creds(worker->creds); > + if (worker->saved_creds) > + revert_creds(worker->saved_creds); > =20 > - worker->creds =3D override_creds(work->creds); > + worker->saved_creds =3D override_creds(work->creds); > + worker->cur_creds =3D work->creds; > } How about as follows? rever_creds() is a bit heavier than put_creds(). static void io_wq_switch_creds(struct io_worker *worker, struct io_wq_work *work) { const struct cred *old_creds =3D override_creds(work->creds); if (worker->saved_creds) put_cred(old_creds); else worker->saved_creds =3D old; worker->cur_creds =3D work->creds; } > =20 > static void io_worker_handle_work(struct io_worker *worker) > @@ -480,7 +482,7 @@ static void io_worker_handle_work(struct io_worker = *worker) > } > if (work->mm !=3D worker->mm) > io_wq_switch_mm(worker, work); > - if (worker->creds !=3D work->creds) > + if (worker->cur_creds !=3D work->creds) > io_wq_switch_creds(worker, work); > /* > * OK to set IO_WQ_WORK_CANCEL even for uncancellable work, >=20 --=20 Pavel Begunkov --cK8FbgRep79RadR9iuAsWyEL8b1jMSHXS-- --lRrc6G6gXrup0csTFFZcEntNplhztBpnq Content-Type: application/pgp-signature; name="signature.asc" Content-Description: OpenPGP digital signature Content-Disposition: attachment; filename="signature.asc" -----BEGIN PGP SIGNATURE----- iQIzBAEBCAAdFiEE+6JuPTjTbx479o3OWt5b1Glr+6UFAl4wzVgACgkQWt5b1Glr +6UP7g//TrctzFamRv2q68uGare3sGHZxaIeAkesfEeJM/ltrvQuh/MIIiWnAdlo 3EE7j5ycRgDpviTaNLd43J3wAGHbz4eZHBeBw9opFkjiAbZird6Etc7+ysLFuojO B8xBhcAfnMN80L4Rb9PTWk6d2pn74YI25mWQxxsbZ2VB44INycF12n1zmeNTF5C7 vOtvr20FVWKE7qDQYbxWbFHhbA0F5btMROOTAqWUtwg0u5sNfgXBst5xKh5yfbjL fj2L3l2aJIhXHzUNFK7pidZw/3/d9RQrpbw6mPcUEdy8wOX7xSaj6KvVRnyLxBx1 wR7Cwm4Kgb5PGEv4lM8NfFR6a6bafB3q0RPsQjyfi6deNoh55PpqdG3P3tUe/7za pHOQ5sm1WIypEgnz2pTPt0D6UxGGji9fI43ihugnB4ju/gLYVf2EG8CUykKUnHFB a83suNfXTKnBR2tub/7Esc2oJ37uX/Um3mYMkXesmempLD+WfSQnaXS+F1mPpD74 hl5LtPhB+tAVExUcZGF78JXogGnUJvbAcdn26nqoq0YPbETMelnGQXWw9+4Ibhik VtZNOqTU9SygfmHjWlchrvtG/6sckWsmxjvl+YFMlRCR6q+NF2GoiRASrRt8dfUA t8kwYMIfclv2iqxyLhbJNqDMklj4/KpZS8nRFBv6SeZbU8zmg9Y= =idCk -----END PGP SIGNATURE----- --lRrc6G6gXrup0csTFFZcEntNplhztBpnq--