From: Julien Grall <julien.grall@arm.com> To: xen-devel@lists.xenproject.org Cc: Oleksandr_Tyshchenko@epam.com, Julien Grall <julien.grall@arm.com>, Stefano Stabellini <sstabellini@kernel.org>, Andrii Anisov <andrii_anisov@epam.com> Subject: [PATCH MM-PART3 v2 10/12] xen/arm: mm: Rework Xen page-tables walk during update Date: Tue, 14 May 2019 13:31:23 +0100 [thread overview] Message-ID: <20190514123125.29086-11-julien.grall@arm.com> (raw) In-Reply-To: <20190514123125.29086-1-julien.grall@arm.com> Currently, xen_pt_update_entry() is only able to update the region covered by xen_second (i.e 0 to 0x7fffffff). Because of the restriction we end to have multiple functions in mm.c modifying the page-tables differently. Furthermore, we never walked the page-tables fully. This means that any change in the layout may requires major rewrite of the page-tables code. Lastly, we have been quite lucky that no one ever tried to pass an address outside this range because it would have blown-up. xen_pt_update_entry() is reworked to walk over the page-tables every time. The logic has been borrowed from arch/arm/p2m.c and contain some limitations for the time being: - Superpage cannot be shattered - Only level 3 (i.e 4KB) can be done Note that the parameter 'addr' has been renamed to 'virt' to make clear we are dealing with a virtual address. Signed-off-by: Julien Grall <julien.grall@arm.com> Reviewed-by: Andrii Anisov <andrii_anisov@epam.com> --- Changes in v2: - Add Andrii's reviewed-by --- xen/arch/arm/mm.c | 121 +++++++++++++++++++++++++++++++++++++++++++++++------- 1 file changed, 106 insertions(+), 15 deletions(-) diff --git a/xen/arch/arm/mm.c b/xen/arch/arm/mm.c index f5979f549b..9a40754f44 100644 --- a/xen/arch/arm/mm.c +++ b/xen/arch/arm/mm.c @@ -984,6 +984,53 @@ static void xen_unmap_table(const lpae_t *table) unmap_domain_page(table); } +#define XEN_TABLE_MAP_FAILED 0 +#define XEN_TABLE_SUPER_PAGE 1 +#define XEN_TABLE_NORMAL_PAGE 2 + +/* + * Take the currently mapped table, find the corresponding entry, + * and map the next table, if available. + * + * The read_only parameters indicates whether intermediate tables should + * be allocated when not present. + * + * Return values: + * XEN_TABLE_MAP_FAILED: Either read_only was set and the entry + * was empty, or allocating a new page failed. + * XEN_TABLE_NORMAL_PAGE: next level mapped normally + * XEN_TABLE_SUPER_PAGE: The next entry points to a superpage. + */ +static int xen_pt_next_level(bool read_only, unsigned int level, + lpae_t **table, unsigned int offset) +{ + lpae_t *entry; + int ret; + + entry = *table + offset; + + if ( !lpae_is_valid(*entry) ) + { + if ( read_only ) + return XEN_TABLE_MAP_FAILED; + + ret = create_xen_table(entry); + if ( ret ) + return XEN_TABLE_MAP_FAILED; + } + + ASSERT(lpae_is_valid(*entry)); + + /* The function xen_pt_next_level is never called at the 3rd level */ + if ( lpae_is_mapping(*entry, level) ) + return XEN_TABLE_SUPER_PAGE; + + xen_unmap_table(*table); + *table = xen_map_table(lpae_get_mfn(*entry)); + + return XEN_TABLE_NORMAL_PAGE; +} + /* Sanity check of the entry */ static bool xen_pt_check_entry(lpae_t entry, mfn_t mfn, unsigned int flags) { @@ -1043,30 +1090,65 @@ static bool xen_pt_check_entry(lpae_t entry, mfn_t mfn, unsigned int flags) return true; } -static int xen_pt_update_entry(unsigned long addr, mfn_t mfn, - unsigned int flags) +static int xen_pt_update_entry(mfn_t root, unsigned long virt, + mfn_t mfn, unsigned int flags) { int rc; + unsigned int level; + /* We only support 4KB mapping (i.e level 3) for now */ + unsigned int target = 3; + lpae_t *table; + /* + * The intermediate page tables are read-only when the MFN is not valid + * and we are not populating page table. + * This means we either modify permissions or remove an entry. + */ + bool read_only = mfn_eq(mfn, INVALID_MFN) && !(flags & _PAGE_POPULATE); lpae_t pte, *entry; - lpae_t *third = NULL; + + /* convenience aliases */ + DECLARE_OFFSETS(offsets, (paddr_t)virt); /* _PAGE_POPULATE and _PAGE_PRESENT should never be set together. */ ASSERT((flags & (_PAGE_POPULATE|_PAGE_PRESENT)) != (_PAGE_POPULATE|_PAGE_PRESENT)); - entry = &xen_second[second_linear_offset(addr)]; - if ( !lpae_is_valid(*entry) || !lpae_is_table(*entry, 2) ) + table = xen_map_table(root); + for ( level = HYP_PT_ROOT_LEVEL; level < target; level++ ) { - int rc = create_xen_table(entry); - if ( rc < 0 ) { - printk("%s: L2 failed\n", __func__); - return rc; + rc = xen_pt_next_level(read_only, level, &table, offsets[level]); + if ( rc == XEN_TABLE_MAP_FAILED ) + { + /* + * We are here because xen_pt_next_level has failed to map + * the intermediate page table (e.g the table does not exist + * and the pt is read-only). It is a valid case when + * removing a mapping as it may not exist in the page table. + * In this case, just ignore it. + */ + if ( flags & (_PAGE_PRESENT|_PAGE_POPULATE) ) + { + mm_printk("%s: Unable to map level %u\n", __func__, level); + rc = -ENOENT; + goto out; + } + else + { + rc = 0; + goto out; + } } + else if ( rc != XEN_TABLE_NORMAL_PAGE ) + break; } - BUG_ON(!lpae_is_valid(*entry)); + if ( level != target ) + { + mm_printk("%s: Shattering superpage is not supported\n", __func__); + rc = -EOPNOTSUPP; + goto out; + } - third = xen_map_table(lpae_get_mfn(*entry)); - entry = &third[third_table_offset(addr)]; + entry = table + offsets[level]; rc = -EINVAL; if ( !xen_pt_check_entry(*entry, mfn, flags) ) @@ -1103,7 +1185,7 @@ static int xen_pt_update_entry(unsigned long addr, mfn_t mfn, rc = 0; out: - xen_unmap_table(third); + xen_unmap_table(table); return rc; } @@ -1119,6 +1201,15 @@ static int xen_pt_update(unsigned long virt, unsigned long addr = virt, addr_end = addr + nr_mfns * PAGE_SIZE; /* + * For arm32, page-tables are different on each CPUs. Yet, they share + * some common mappings. It is assumed that only common mappings + * will be modified with this function. + * + * XXX: Add a check. + */ + const mfn_t root = virt_to_mfn(THIS_CPU_PGTABLE); + + /* * The hardware was configured to forbid mapping both writeable and * executable. * When modifying/creating mapping (i.e _PAGE_PRESENT is set), @@ -1139,9 +1230,9 @@ static int xen_pt_update(unsigned long virt, spin_lock(&xen_pt_lock); - for( ; addr < addr_end; addr += PAGE_SIZE ) + for ( ; addr < addr_end; addr += PAGE_SIZE ) { - rc = xen_pt_update_entry(addr, mfn, flags); + rc = xen_pt_update_entry(root, addr, mfn, flags); if ( rc ) break; -- 2.11.0 _______________________________________________ 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: Julien Grall <julien.grall@arm.com> To: xen-devel@lists.xenproject.org Cc: Oleksandr_Tyshchenko@epam.com, Julien Grall <julien.grall@arm.com>, Stefano Stabellini <sstabellini@kernel.org>, Andrii Anisov <andrii_anisov@epam.com> Subject: [Xen-devel] [PATCH MM-PART3 v2 10/12] xen/arm: mm: Rework Xen page-tables walk during update Date: Tue, 14 May 2019 13:31:23 +0100 [thread overview] Message-ID: <20190514123125.29086-11-julien.grall@arm.com> (raw) Message-ID: <20190514123123.VSM1dp_JTTMJo6CcmJKR1raQUhOcDmwARR7_hemdXmo@z> (raw) In-Reply-To: <20190514123125.29086-1-julien.grall@arm.com> Currently, xen_pt_update_entry() is only able to update the region covered by xen_second (i.e 0 to 0x7fffffff). Because of the restriction we end to have multiple functions in mm.c modifying the page-tables differently. Furthermore, we never walked the page-tables fully. This means that any change in the layout may requires major rewrite of the page-tables code. Lastly, we have been quite lucky that no one ever tried to pass an address outside this range because it would have blown-up. xen_pt_update_entry() is reworked to walk over the page-tables every time. The logic has been borrowed from arch/arm/p2m.c and contain some limitations for the time being: - Superpage cannot be shattered - Only level 3 (i.e 4KB) can be done Note that the parameter 'addr' has been renamed to 'virt' to make clear we are dealing with a virtual address. Signed-off-by: Julien Grall <julien.grall@arm.com> Reviewed-by: Andrii Anisov <andrii_anisov@epam.com> --- Changes in v2: - Add Andrii's reviewed-by --- xen/arch/arm/mm.c | 121 +++++++++++++++++++++++++++++++++++++++++++++++------- 1 file changed, 106 insertions(+), 15 deletions(-) diff --git a/xen/arch/arm/mm.c b/xen/arch/arm/mm.c index f5979f549b..9a40754f44 100644 --- a/xen/arch/arm/mm.c +++ b/xen/arch/arm/mm.c @@ -984,6 +984,53 @@ static void xen_unmap_table(const lpae_t *table) unmap_domain_page(table); } +#define XEN_TABLE_MAP_FAILED 0 +#define XEN_TABLE_SUPER_PAGE 1 +#define XEN_TABLE_NORMAL_PAGE 2 + +/* + * Take the currently mapped table, find the corresponding entry, + * and map the next table, if available. + * + * The read_only parameters indicates whether intermediate tables should + * be allocated when not present. + * + * Return values: + * XEN_TABLE_MAP_FAILED: Either read_only was set and the entry + * was empty, or allocating a new page failed. + * XEN_TABLE_NORMAL_PAGE: next level mapped normally + * XEN_TABLE_SUPER_PAGE: The next entry points to a superpage. + */ +static int xen_pt_next_level(bool read_only, unsigned int level, + lpae_t **table, unsigned int offset) +{ + lpae_t *entry; + int ret; + + entry = *table + offset; + + if ( !lpae_is_valid(*entry) ) + { + if ( read_only ) + return XEN_TABLE_MAP_FAILED; + + ret = create_xen_table(entry); + if ( ret ) + return XEN_TABLE_MAP_FAILED; + } + + ASSERT(lpae_is_valid(*entry)); + + /* The function xen_pt_next_level is never called at the 3rd level */ + if ( lpae_is_mapping(*entry, level) ) + return XEN_TABLE_SUPER_PAGE; + + xen_unmap_table(*table); + *table = xen_map_table(lpae_get_mfn(*entry)); + + return XEN_TABLE_NORMAL_PAGE; +} + /* Sanity check of the entry */ static bool xen_pt_check_entry(lpae_t entry, mfn_t mfn, unsigned int flags) { @@ -1043,30 +1090,65 @@ static bool xen_pt_check_entry(lpae_t entry, mfn_t mfn, unsigned int flags) return true; } -static int xen_pt_update_entry(unsigned long addr, mfn_t mfn, - unsigned int flags) +static int xen_pt_update_entry(mfn_t root, unsigned long virt, + mfn_t mfn, unsigned int flags) { int rc; + unsigned int level; + /* We only support 4KB mapping (i.e level 3) for now */ + unsigned int target = 3; + lpae_t *table; + /* + * The intermediate page tables are read-only when the MFN is not valid + * and we are not populating page table. + * This means we either modify permissions or remove an entry. + */ + bool read_only = mfn_eq(mfn, INVALID_MFN) && !(flags & _PAGE_POPULATE); lpae_t pte, *entry; - lpae_t *third = NULL; + + /* convenience aliases */ + DECLARE_OFFSETS(offsets, (paddr_t)virt); /* _PAGE_POPULATE and _PAGE_PRESENT should never be set together. */ ASSERT((flags & (_PAGE_POPULATE|_PAGE_PRESENT)) != (_PAGE_POPULATE|_PAGE_PRESENT)); - entry = &xen_second[second_linear_offset(addr)]; - if ( !lpae_is_valid(*entry) || !lpae_is_table(*entry, 2) ) + table = xen_map_table(root); + for ( level = HYP_PT_ROOT_LEVEL; level < target; level++ ) { - int rc = create_xen_table(entry); - if ( rc < 0 ) { - printk("%s: L2 failed\n", __func__); - return rc; + rc = xen_pt_next_level(read_only, level, &table, offsets[level]); + if ( rc == XEN_TABLE_MAP_FAILED ) + { + /* + * We are here because xen_pt_next_level has failed to map + * the intermediate page table (e.g the table does not exist + * and the pt is read-only). It is a valid case when + * removing a mapping as it may not exist in the page table. + * In this case, just ignore it. + */ + if ( flags & (_PAGE_PRESENT|_PAGE_POPULATE) ) + { + mm_printk("%s: Unable to map level %u\n", __func__, level); + rc = -ENOENT; + goto out; + } + else + { + rc = 0; + goto out; + } } + else if ( rc != XEN_TABLE_NORMAL_PAGE ) + break; } - BUG_ON(!lpae_is_valid(*entry)); + if ( level != target ) + { + mm_printk("%s: Shattering superpage is not supported\n", __func__); + rc = -EOPNOTSUPP; + goto out; + } - third = xen_map_table(lpae_get_mfn(*entry)); - entry = &third[third_table_offset(addr)]; + entry = table + offsets[level]; rc = -EINVAL; if ( !xen_pt_check_entry(*entry, mfn, flags) ) @@ -1103,7 +1185,7 @@ static int xen_pt_update_entry(unsigned long addr, mfn_t mfn, rc = 0; out: - xen_unmap_table(third); + xen_unmap_table(table); return rc; } @@ -1119,6 +1201,15 @@ static int xen_pt_update(unsigned long virt, unsigned long addr = virt, addr_end = addr + nr_mfns * PAGE_SIZE; /* + * For arm32, page-tables are different on each CPUs. Yet, they share + * some common mappings. It is assumed that only common mappings + * will be modified with this function. + * + * XXX: Add a check. + */ + const mfn_t root = virt_to_mfn(THIS_CPU_PGTABLE); + + /* * The hardware was configured to forbid mapping both writeable and * executable. * When modifying/creating mapping (i.e _PAGE_PRESENT is set), @@ -1139,9 +1230,9 @@ static int xen_pt_update(unsigned long virt, spin_lock(&xen_pt_lock); - for( ; addr < addr_end; addr += PAGE_SIZE ) + for ( ; addr < addr_end; addr += PAGE_SIZE ) { - rc = xen_pt_update_entry(addr, mfn, flags); + rc = xen_pt_update_entry(root, addr, mfn, flags); if ( rc ) break; -- 2.11.0 _______________________________________________ Xen-devel mailing list Xen-devel@lists.xenproject.org https://lists.xenproject.org/mailman/listinfo/xen-devel
next prev parent reply other threads:[~2019-05-14 12:31 UTC|newest] Thread overview: 64+ messages / expand[flat|nested] mbox.gz Atom feed top 2019-05-14 12:31 [PATCH MM-PART3 v2 00/12] xen/arm: Provide a generic function to update Xen PT Julien Grall 2019-05-14 12:31 ` [Xen-devel] " Julien Grall 2019-05-14 12:31 ` [PATCH MM-PART3 v2 01/12] xen/arm: lpae: Add a macro to generate offsets from an address Julien Grall 2019-05-14 12:31 ` [Xen-devel] " Julien Grall 2019-06-11 18:21 ` Stefano Stabellini 2019-06-11 18:27 ` Julien Grall 2019-05-14 12:31 ` [PATCH MM-PART3 v2 02/12] xen/arm: mm: Rename create_xen_entries() to xen_pt_update() Julien Grall 2019-05-14 12:31 ` [Xen-devel] " Julien Grall 2019-06-11 18:23 ` Stefano Stabellini 2019-05-14 12:31 ` [PATCH MM-PART3 v2 03/12] xen/arm: mm: Move out of xen_pt_update() the logic to update an entry Julien Grall 2019-05-14 12:31 ` [Xen-devel] " Julien Grall 2019-06-11 18:29 ` Stefano Stabellini 2019-05-14 12:31 ` [PATCH MM-PART3 v2 04/12] xen/arm: mm: Only increment mfn when valid in xen_pt_update Julien Grall 2019-05-14 12:31 ` [Xen-devel] " Julien Grall 2019-06-11 18:37 ` Stefano Stabellini 2019-06-11 19:56 ` [Xen-devel] Checking INVALID_MFN in mfn_add (WAS: Re: [PATCH MM-PART3 v2 04/12] xen/arm: mm: Only increment mfn when valid in xen_pt_update) Julien Grall 2019-06-11 20:24 ` Andrew Cooper 2019-06-12 12:47 ` Julien Grall 2019-06-12 15:57 ` Stefano Stabellini 2019-06-12 7:53 ` Jan Beulich 2019-05-14 12:31 ` [PATCH MM-PART3 v2 05/12] xen/arm: mm: Introduce _PAGE_PRESENT and _PAGE_POPULATE Julien Grall 2019-05-14 12:31 ` [Xen-devel] " Julien Grall 2019-06-11 22:35 ` Stefano Stabellini 2019-06-12 13:00 ` Julien Grall 2019-05-14 12:31 ` [PATCH MM-PART3 v2 06/12] xen/arm: mm: Sanity check any update of Xen page tables Julien Grall 2019-05-14 12:31 ` [Xen-devel] " Julien Grall 2019-06-12 0:10 ` Stefano Stabellini 2019-06-12 14:48 ` Julien Grall 2019-06-12 15:54 ` Stefano Stabellini 2019-06-12 15:58 ` Julien Grall 2019-05-14 12:31 ` [PATCH MM-PART3 v2 07/12] xen/arm: mm: Rework xen_pt_update_entry to avoid use xenmap_operation Julien Grall 2019-05-14 12:31 ` [Xen-devel] " Julien Grall 2019-06-12 22:22 ` Stefano Stabellini 2019-05-14 12:31 ` [PATCH MM-PART3 v2 08/12] xen/arm: mm: Remove enum xenmap_operation Julien Grall 2019-05-14 12:31 ` [Xen-devel] " Julien Grall 2019-06-11 22:38 ` Stefano Stabellini 2019-05-14 12:31 ` [PATCH MM-PART3 v2 09/12] xen/arm: mm: Use {, un}map_domain_page() to map/unmap Xen page-tables Julien Grall 2019-05-14 12:31 ` [Xen-devel] " Julien Grall 2019-06-12 22:25 ` Stefano Stabellini 2019-06-13 8:07 ` Julien Grall 2019-06-13 17:55 ` Stefano Stabellini 2019-05-14 12:31 ` Julien Grall [this message] 2019-05-14 12:31 ` [Xen-devel] [PATCH MM-PART3 v2 10/12] xen/arm: mm: Rework Xen page-tables walk during update Julien Grall 2019-06-12 22:52 ` Stefano Stabellini 2019-06-13 8:20 ` Julien Grall 2019-06-13 17:59 ` Stefano Stabellini 2019-06-13 21:32 ` Julien Grall 2019-06-13 22:57 ` Stefano Stabellini 2019-05-14 12:31 ` [PATCH MM-PART3 v2 11/12] xen/arm: mm: Don't open-code Xen PT update in {set, clear}_fixmap() Julien Grall 2019-05-14 12:31 ` [Xen-devel] " Julien Grall 2019-06-12 22:33 ` Stefano Stabellini 2019-06-13 8:31 ` Julien Grall 2019-06-13 18:51 ` Stefano Stabellini 2019-06-13 21:21 ` Julien Grall 2019-06-13 22:55 ` Stefano Stabellini 2019-05-14 12:31 ` [PATCH MM-PART3 v2 12/12] xen/arm: mm: Remove set_pte_flags_on_range() Julien Grall 2019-05-14 12:31 ` [Xen-devel] " Julien Grall 2019-06-12 22:41 ` Stefano Stabellini 2019-06-13 8:51 ` Julien Grall 2019-06-13 18:04 ` Stefano Stabellini 2019-06-13 21:22 ` Julien Grall 2019-05-29 17:23 ` [PATCH MM-PART3 v2 00/12] xen/arm: Provide a generic function to update Xen PT Julien Grall 2019-05-29 17:23 ` [Xen-devel] " Julien Grall 2019-06-10 10:08 ` Julien Grall
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=20190514123125.29086-11-julien.grall@arm.com \ --to=julien.grall@arm.com \ --cc=Oleksandr_Tyshchenko@epam.com \ --cc=andrii_anisov@epam.com \ --cc=sstabellini@kernel.org \ --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.