From: Tamas K Lengyel <tamas@tklengyel.com> To: Alexandru Stefan ISAILA <aisaila@bitdefender.com> Cc: "wei.liu2@citrix.com" <wei.liu2@citrix.com>, "rcojocaru@bitdefender.com" <rcojocaru@bitdefender.com>, "george.dunlap@eu.citrix.com" <george.dunlap@eu.citrix.com>, "andrew.cooper3@citrix.com" <andrew.cooper3@citrix.com>, "jbeulich@suse.com" <jbeulich@suse.com>, "xen-devel@lists.xenproject.org" <xen-devel@lists.xenproject.org>, "roger.pau@citrix.com" <roger.pau@citrix.com> Subject: Re: [PATCH v3 1/3] x86/mm: Introduce altp2m_get_gfn_type_access Date: Tue, 9 Apr 2019 07:48:56 -0600 [thread overview] Message-ID: <CABfawhmyTGNmLP5=c5pWeUiRaSPpwUUw_9SjXoJPr2ZANep41w@mail.gmail.com> (raw) In-Reply-To: <20190409120324.13940-1-aisaila@bitdefender.com> On Tue, Apr 9, 2019 at 6:04 AM Alexandru Stefan ISAILA <aisaila@bitdefender.com> wrote: > > This patch moves common code from p2m_set_altp2m_mem_access() and > p2m_change_altp2m_gfn() into one function > > Signed-off-by: Alexandru Isaila <aisaila@bitdefender.com> > > --- > Changes since V2: > - Change var name from found_in_hostp2m to copied_from_hostp2m > - Move the type check from altp2m_get_gfn_type_access() to the > callers. > --- > xen/arch/x86/mm/mem_access.c | 32 ++++++++++++---------------- > xen/arch/x86/mm/p2m.c | 41 ++++++++++++++---------------------- > xen/include/asm-x86/p2m.h | 19 +++++++++++++++++ > 3 files changed, 49 insertions(+), 43 deletions(-) > > diff --git a/xen/arch/x86/mm/mem_access.c b/xen/arch/x86/mm/mem_access.c > index 56c06a4fc6..bf67ddb15a 100644 > --- a/xen/arch/x86/mm/mem_access.c > +++ b/xen/arch/x86/mm/mem_access.c > @@ -265,31 +265,27 @@ int p2m_set_altp2m_mem_access(struct domain *d, struct p2m_domain *hp2m, > unsigned int page_order; > unsigned long gfn_l = gfn_x(gfn); > int rc; > + bool copied_from_hostp2m; > > - mfn = ap2m->get_entry(ap2m, gfn, &t, &old_a, 0, NULL, NULL); > + mfn = altp2m_get_gfn_type_access(ap2m, gfn, &t, &old_a, &page_order, &copied_from_hostp2m); > > - /* Check host p2m if no valid entry in alternate */ > if ( !mfn_valid(mfn) ) > + return -ESRCH; > + > + /* If this is a superpage, copy that first */ > + if ( page_order != PAGE_ORDER_4K && copied_from_hostp2m ) > { > + unsigned long mask = ~((1UL << page_order) - 1); > + gfn_t gfn2 = _gfn(gfn_l & mask); > + mfn_t mfn2 = _mfn(mfn_x(mfn) & mask); > > - mfn = __get_gfn_type_access(hp2m, gfn_l, &t, &old_a, > - P2M_ALLOC | P2M_UNSHARE, &page_order, 0); > + /* Note: currently it is not safe to remap to a shared entry */ > + if ( t != p2m_ram_rw ) > + return -ESRCH; > > - rc = -ESRCH; > - if ( !mfn_valid(mfn) || t != p2m_ram_rw ) > + rc = ap2m->set_entry(ap2m, gfn2, mfn2, page_order, t, old_a, 1); > + if ( rc ) > return rc; > - > - /* If this is a superpage, copy that first */ > - if ( page_order != PAGE_ORDER_4K ) > - { > - unsigned long mask = ~((1UL << page_order) - 1); > - gfn_t gfn2 = _gfn(gfn_l & mask); > - mfn_t mfn2 = _mfn(mfn_x(mfn) & mask); > - > - rc = ap2m->set_entry(ap2m, gfn2, mfn2, page_order, t, old_a, 1); > - if ( rc ) > - return rc; > - } > } > > /* > diff --git a/xen/arch/x86/mm/p2m.c b/xen/arch/x86/mm/p2m.c > index b9bbb8f485..d38d7c29ca 100644 > --- a/xen/arch/x86/mm/p2m.c > +++ b/xen/arch/x86/mm/p2m.c > @@ -2626,6 +2626,7 @@ int p2m_change_altp2m_gfn(struct domain *d, unsigned int idx, > mfn_t mfn; > unsigned int page_order; > int rc = -EINVAL; > + bool copied_from_hostp2m; > > if ( idx >= MAX_ALTP2M || d->arch.altp2m_eptp[idx] == mfn_x(INVALID_MFN) ) > return rc; > @@ -2636,7 +2637,7 @@ int p2m_change_altp2m_gfn(struct domain *d, unsigned int idx, > p2m_lock(hp2m); > p2m_lock(ap2m); > > - mfn = ap2m->get_entry(ap2m, old_gfn, &t, &a, 0, NULL, NULL); > + mfn = altp2m_get_gfn_type_access(ap2m, old_gfn, &t, &a, &page_order, &copied_from_hostp2m); > > if ( gfn_eq(new_gfn, INVALID_GFN) ) > { > @@ -2646,37 +2647,27 @@ int p2m_change_altp2m_gfn(struct domain *d, unsigned int idx, > goto out; > } > > - /* Check host p2m if no valid entry in alternate */ > - if ( !mfn_valid(mfn) ) > - { > - mfn = __get_gfn_type_access(hp2m, gfn_x(old_gfn), &t, &a, > - P2M_ALLOC, &page_order, 0); > + if ( !mfn_valid(mfn) || (t != p2m_ram_rw && copied_from_hostp2m) ) Is this check correct? Why do you want to get out only when type is non-rw *and* it's copied from the hostp2m? You could have non-rw entries like mmio in the altp2m that were lazily copied and I don't think we want to allow remapping to those either. > + goto out; > > - if ( !mfn_valid(mfn) || t != p2m_ram_rw ) > - goto out; > - > - /* If this is a superpage, copy that first */ > - if ( page_order != PAGE_ORDER_4K ) > - { > - gfn_t gfn; > - unsigned long mask; > + /* If this is a superpage, copy that first */ > + if ( page_order != PAGE_ORDER_4K && copied_from_hostp2m ) > + { > + gfn_t gfn; > + unsigned long mask; > > - mask = ~((1UL << page_order) - 1); > - gfn = _gfn(gfn_x(old_gfn) & mask); > - mfn = _mfn(mfn_x(mfn) & mask); > + mask = ~((1UL << page_order) - 1); > + gfn = _gfn(gfn_x(old_gfn) & mask); > + mfn = _mfn(mfn_x(mfn) & mask); > > - if ( ap2m->set_entry(ap2m, gfn, mfn, page_order, t, a, 1) ) > - goto out; > - } > + if ( ap2m->set_entry(ap2m, gfn, mfn, page_order, t, a, 1) ) > + goto out; > } > > - mfn = ap2m->get_entry(ap2m, new_gfn, &t, &a, 0, NULL, NULL); > - > - if ( !mfn_valid(mfn) ) > - mfn = hp2m->get_entry(hp2m, new_gfn, &t, &a, 0, NULL, NULL); > + mfn = altp2m_get_gfn_type_access(ap2m, new_gfn, &t, &a, &page_order, &copied_from_hostp2m); > > /* Note: currently it is not safe to remap to a shared entry */ > - if ( !mfn_valid(mfn) || (t != p2m_ram_rw) ) > + if ( !mfn_valid(mfn) || (t != p2m_ram_rw && copied_from_hostp2m) ) > goto out; > > if ( !ap2m->set_entry(ap2m, old_gfn, mfn, PAGE_ORDER_4K, t, a, > diff --git a/xen/include/asm-x86/p2m.h b/xen/include/asm-x86/p2m.h > index 2801a8ccca..6de1546d76 100644 > --- a/xen/include/asm-x86/p2m.h > +++ b/xen/include/asm-x86/p2m.h > @@ -448,6 +448,25 @@ static inline mfn_t __nonnull(3) get_gfn_type( > return get_gfn_type_access(p2m_get_hostp2m(d), gfn, t, &a, q, NULL); > } > > +static inline mfn_t altp2m_get_gfn_type_access( > + struct p2m_domain *ap2m, gfn_t gfn, p2m_type_t *t, p2m_access_t *a, > + unsigned int *page_order, bool *copied_from_hostp2m) > +{ > + mfn_t mfn = ap2m->get_entry(ap2m, gfn, t, a, 0, NULL, NULL); > + > + *copied_from_hostp2m = false; > + > + /* Check host p2m if no valid entry in alternate */ > + if ( !mfn_valid(mfn) ) > + { > + mfn = __get_gfn_type_access(p2m_get_hostp2m(ap2m->domain), gfn_x(gfn), t, a, > + P2M_ALLOC | P2M_UNSHARE, page_order, false); > + *copied_from_hostp2m = mfn_valid(mfn); > + } > + > + return mfn; > +} > + > /* Syntactic sugar: most callers will use one of these. */ > #define get_gfn(d, g, t) get_gfn_type((d), (g), (t), P2M_ALLOC) > #define get_gfn_query(d, g, t) get_gfn_type((d), (g), (t), 0) > -- > 2.17.1 > _______________________________________________ Xen-devel mailing list Xen-devel@lists.xenproject.org https://lists.xenproject.org/mailman/listinfo/xen-devel
WARNING: multiple messages have this Message-ID (diff)
From: Tamas K Lengyel <tamas@tklengyel.com> To: Alexandru Stefan ISAILA <aisaila@bitdefender.com> Cc: "wei.liu2@citrix.com" <wei.liu2@citrix.com>, "rcojocaru@bitdefender.com" <rcojocaru@bitdefender.com>, "george.dunlap@eu.citrix.com" <george.dunlap@eu.citrix.com>, "andrew.cooper3@citrix.com" <andrew.cooper3@citrix.com>, "jbeulich@suse.com" <jbeulich@suse.com>, "xen-devel@lists.xenproject.org" <xen-devel@lists.xenproject.org>, "roger.pau@citrix.com" <roger.pau@citrix.com> Subject: Re: [Xen-devel] [PATCH v3 1/3] x86/mm: Introduce altp2m_get_gfn_type_access Date: Tue, 9 Apr 2019 07:48:56 -0600 [thread overview] Message-ID: <CABfawhmyTGNmLP5=c5pWeUiRaSPpwUUw_9SjXoJPr2ZANep41w@mail.gmail.com> (raw) Message-ID: <20190409134856.V5ChDDSB1WF6lIQvMi14KxI6xtlj1YtGQ6bD-fRWfGI@z> (raw) In-Reply-To: <20190409120324.13940-1-aisaila@bitdefender.com> On Tue, Apr 9, 2019 at 6:04 AM Alexandru Stefan ISAILA <aisaila@bitdefender.com> wrote: > > This patch moves common code from p2m_set_altp2m_mem_access() and > p2m_change_altp2m_gfn() into one function > > Signed-off-by: Alexandru Isaila <aisaila@bitdefender.com> > > --- > Changes since V2: > - Change var name from found_in_hostp2m to copied_from_hostp2m > - Move the type check from altp2m_get_gfn_type_access() to the > callers. > --- > xen/arch/x86/mm/mem_access.c | 32 ++++++++++++---------------- > xen/arch/x86/mm/p2m.c | 41 ++++++++++++++---------------------- > xen/include/asm-x86/p2m.h | 19 +++++++++++++++++ > 3 files changed, 49 insertions(+), 43 deletions(-) > > diff --git a/xen/arch/x86/mm/mem_access.c b/xen/arch/x86/mm/mem_access.c > index 56c06a4fc6..bf67ddb15a 100644 > --- a/xen/arch/x86/mm/mem_access.c > +++ b/xen/arch/x86/mm/mem_access.c > @@ -265,31 +265,27 @@ int p2m_set_altp2m_mem_access(struct domain *d, struct p2m_domain *hp2m, > unsigned int page_order; > unsigned long gfn_l = gfn_x(gfn); > int rc; > + bool copied_from_hostp2m; > > - mfn = ap2m->get_entry(ap2m, gfn, &t, &old_a, 0, NULL, NULL); > + mfn = altp2m_get_gfn_type_access(ap2m, gfn, &t, &old_a, &page_order, &copied_from_hostp2m); > > - /* Check host p2m if no valid entry in alternate */ > if ( !mfn_valid(mfn) ) > + return -ESRCH; > + > + /* If this is a superpage, copy that first */ > + if ( page_order != PAGE_ORDER_4K && copied_from_hostp2m ) > { > + unsigned long mask = ~((1UL << page_order) - 1); > + gfn_t gfn2 = _gfn(gfn_l & mask); > + mfn_t mfn2 = _mfn(mfn_x(mfn) & mask); > > - mfn = __get_gfn_type_access(hp2m, gfn_l, &t, &old_a, > - P2M_ALLOC | P2M_UNSHARE, &page_order, 0); > + /* Note: currently it is not safe to remap to a shared entry */ > + if ( t != p2m_ram_rw ) > + return -ESRCH; > > - rc = -ESRCH; > - if ( !mfn_valid(mfn) || t != p2m_ram_rw ) > + rc = ap2m->set_entry(ap2m, gfn2, mfn2, page_order, t, old_a, 1); > + if ( rc ) > return rc; > - > - /* If this is a superpage, copy that first */ > - if ( page_order != PAGE_ORDER_4K ) > - { > - unsigned long mask = ~((1UL << page_order) - 1); > - gfn_t gfn2 = _gfn(gfn_l & mask); > - mfn_t mfn2 = _mfn(mfn_x(mfn) & mask); > - > - rc = ap2m->set_entry(ap2m, gfn2, mfn2, page_order, t, old_a, 1); > - if ( rc ) > - return rc; > - } > } > > /* > diff --git a/xen/arch/x86/mm/p2m.c b/xen/arch/x86/mm/p2m.c > index b9bbb8f485..d38d7c29ca 100644 > --- a/xen/arch/x86/mm/p2m.c > +++ b/xen/arch/x86/mm/p2m.c > @@ -2626,6 +2626,7 @@ int p2m_change_altp2m_gfn(struct domain *d, unsigned int idx, > mfn_t mfn; > unsigned int page_order; > int rc = -EINVAL; > + bool copied_from_hostp2m; > > if ( idx >= MAX_ALTP2M || d->arch.altp2m_eptp[idx] == mfn_x(INVALID_MFN) ) > return rc; > @@ -2636,7 +2637,7 @@ int p2m_change_altp2m_gfn(struct domain *d, unsigned int idx, > p2m_lock(hp2m); > p2m_lock(ap2m); > > - mfn = ap2m->get_entry(ap2m, old_gfn, &t, &a, 0, NULL, NULL); > + mfn = altp2m_get_gfn_type_access(ap2m, old_gfn, &t, &a, &page_order, &copied_from_hostp2m); > > if ( gfn_eq(new_gfn, INVALID_GFN) ) > { > @@ -2646,37 +2647,27 @@ int p2m_change_altp2m_gfn(struct domain *d, unsigned int idx, > goto out; > } > > - /* Check host p2m if no valid entry in alternate */ > - if ( !mfn_valid(mfn) ) > - { > - mfn = __get_gfn_type_access(hp2m, gfn_x(old_gfn), &t, &a, > - P2M_ALLOC, &page_order, 0); > + if ( !mfn_valid(mfn) || (t != p2m_ram_rw && copied_from_hostp2m) ) Is this check correct? Why do you want to get out only when type is non-rw *and* it's copied from the hostp2m? You could have non-rw entries like mmio in the altp2m that were lazily copied and I don't think we want to allow remapping to those either. > + goto out; > > - if ( !mfn_valid(mfn) || t != p2m_ram_rw ) > - goto out; > - > - /* If this is a superpage, copy that first */ > - if ( page_order != PAGE_ORDER_4K ) > - { > - gfn_t gfn; > - unsigned long mask; > + /* If this is a superpage, copy that first */ > + if ( page_order != PAGE_ORDER_4K && copied_from_hostp2m ) > + { > + gfn_t gfn; > + unsigned long mask; > > - mask = ~((1UL << page_order) - 1); > - gfn = _gfn(gfn_x(old_gfn) & mask); > - mfn = _mfn(mfn_x(mfn) & mask); > + mask = ~((1UL << page_order) - 1); > + gfn = _gfn(gfn_x(old_gfn) & mask); > + mfn = _mfn(mfn_x(mfn) & mask); > > - if ( ap2m->set_entry(ap2m, gfn, mfn, page_order, t, a, 1) ) > - goto out; > - } > + if ( ap2m->set_entry(ap2m, gfn, mfn, page_order, t, a, 1) ) > + goto out; > } > > - mfn = ap2m->get_entry(ap2m, new_gfn, &t, &a, 0, NULL, NULL); > - > - if ( !mfn_valid(mfn) ) > - mfn = hp2m->get_entry(hp2m, new_gfn, &t, &a, 0, NULL, NULL); > + mfn = altp2m_get_gfn_type_access(ap2m, new_gfn, &t, &a, &page_order, &copied_from_hostp2m); > > /* Note: currently it is not safe to remap to a shared entry */ > - if ( !mfn_valid(mfn) || (t != p2m_ram_rw) ) > + if ( !mfn_valid(mfn) || (t != p2m_ram_rw && copied_from_hostp2m) ) > goto out; > > if ( !ap2m->set_entry(ap2m, old_gfn, mfn, PAGE_ORDER_4K, t, a, > diff --git a/xen/include/asm-x86/p2m.h b/xen/include/asm-x86/p2m.h > index 2801a8ccca..6de1546d76 100644 > --- a/xen/include/asm-x86/p2m.h > +++ b/xen/include/asm-x86/p2m.h > @@ -448,6 +448,25 @@ static inline mfn_t __nonnull(3) get_gfn_type( > return get_gfn_type_access(p2m_get_hostp2m(d), gfn, t, &a, q, NULL); > } > > +static inline mfn_t altp2m_get_gfn_type_access( > + struct p2m_domain *ap2m, gfn_t gfn, p2m_type_t *t, p2m_access_t *a, > + unsigned int *page_order, bool *copied_from_hostp2m) > +{ > + mfn_t mfn = ap2m->get_entry(ap2m, gfn, t, a, 0, NULL, NULL); > + > + *copied_from_hostp2m = false; > + > + /* Check host p2m if no valid entry in alternate */ > + if ( !mfn_valid(mfn) ) > + { > + mfn = __get_gfn_type_access(p2m_get_hostp2m(ap2m->domain), gfn_x(gfn), t, a, > + P2M_ALLOC | P2M_UNSHARE, page_order, false); > + *copied_from_hostp2m = mfn_valid(mfn); > + } > + > + return mfn; > +} > + > /* Syntactic sugar: most callers will use one of these. */ > #define get_gfn(d, g, t) get_gfn_type((d), (g), (t), P2M_ALLOC) > #define get_gfn_query(d, g, t) get_gfn_type((d), (g), (t), 0) > -- > 2.17.1 > _______________________________________________ Xen-devel mailing list Xen-devel@lists.xenproject.org https://lists.xenproject.org/mailman/listinfo/xen-devel
next prev parent reply other threads:[~2019-04-09 13:49 UTC|newest] Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top 2019-04-09 12:03 [PATCH v3 1/3] x86/mm: Introduce altp2m_get_gfn_type_access Alexandru Stefan ISAILA 2019-04-09 12:03 ` [Xen-devel] " Alexandru Stefan ISAILA 2019-04-09 12:03 ` [PATCH v3 2/3] x86/mm: Introduce altp2m_set_entry_by_page_order Alexandru Stefan ISAILA 2019-04-09 12:03 ` [Xen-devel] " Alexandru Stefan ISAILA 2019-04-09 13:44 ` Tamas K Lengyel 2019-04-09 13:44 ` [Xen-devel] " Tamas K Lengyel 2019-04-10 14:18 ` George Dunlap 2019-04-10 14:18 ` [Xen-devel] " George Dunlap 2019-04-10 14:22 ` Alexandru Stefan ISAILA 2019-04-10 14:22 ` [Xen-devel] " Alexandru Stefan ISAILA 2019-05-02 14:46 ` Jan Beulich 2019-05-02 14:46 ` [Xen-devel] " Jan Beulich 2019-04-09 12:04 ` [PATCH v3 3/3] x86/mm: Fix p2m_set_suppress_ve Alexandru Stefan ISAILA 2019-04-09 12:04 ` [Xen-devel] " Alexandru Stefan ISAILA 2019-04-09 13:48 ` Tamas K Lengyel [this message] 2019-04-09 13:48 ` [Xen-devel] [PATCH v3 1/3] x86/mm: Introduce altp2m_get_gfn_type_access Tamas K Lengyel 2019-04-09 14:03 ` Alexandru Stefan ISAILA 2019-04-09 14:03 ` [Xen-devel] " Alexandru Stefan ISAILA 2019-04-09 14:37 ` Tamas K Lengyel 2019-04-09 14:37 ` [Xen-devel] " Tamas K Lengyel 2019-04-09 14:48 ` Alexandru Stefan ISAILA 2019-04-09 14:48 ` [Xen-devel] " Alexandru Stefan ISAILA 2019-04-09 15:26 ` Tamas K Lengyel 2019-04-09 15:26 ` [Xen-devel] " Tamas K Lengyel 2019-04-09 15:37 ` Alexandru Stefan ISAILA 2019-04-09 15:37 ` [Xen-devel] " Alexandru Stefan ISAILA 2019-04-10 16:02 ` George Dunlap 2019-04-10 16:02 ` [Xen-devel] " George Dunlap 2019-04-11 12:17 ` Alexandru Stefan ISAILA 2019-04-11 12:17 ` [Xen-devel] " Alexandru Stefan ISAILA 2019-04-11 12:50 ` George Dunlap 2019-04-11 12:50 ` [Xen-devel] " George Dunlap 2019-04-11 13:28 ` Tamas K Lengyel 2019-04-11 13:28 ` [Xen-devel] " Tamas K Lengyel 2019-04-12 10:59 ` Alexandru Stefan ISAILA 2019-04-12 10:59 ` [Xen-devel] " Alexandru Stefan ISAILA 2019-05-02 14:43 ` Jan Beulich 2019-05-02 14:43 ` [Xen-devel] " Jan Beulich
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to='CABfawhmyTGNmLP5=c5pWeUiRaSPpwUUw_9SjXoJPr2ZANep41w@mail.gmail.com' \ --to=tamas@tklengyel.com \ --cc=aisaila@bitdefender.com \ --cc=andrew.cooper3@citrix.com \ --cc=george.dunlap@eu.citrix.com \ --cc=jbeulich@suse.com \ --cc=rcojocaru@bitdefender.com \ --cc=roger.pau@citrix.com \ --cc=wei.liu2@citrix.com \ --cc=xen-devel@lists.xenproject.org \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.