xen: fix dom0 PV boot on some AMD machines
Subsystems (3)
Juergen Gross
maintainerv2 cover xen: fix dom0 PV boot on some AMD machines
There have been reports of failed boots with Xen due to an overlap of the kernel's memory with ACPI NVS reported in the E820 map of the host.
This series fixes this issue by moving the NVS area in dom0 to some higher address.
Changes in V2:
- split of v1 patch 5
- new patch 6
Juergen Gross (7):
xen: use correct end address of kernel for conflict checking
xen: introduce generic helper checking for memory map conflicts
xen: move checks for e820 conflicts further up
xen: move max_pfn in xen_memory_setup() out of function scope
xen: add capability to remap non-RAM pages to different PFNs
xen: allow mapping ACPI data using a different physical address
xen: tolerate ACPI NVS memory overlapping with Xen allocated memory
arch/x86/include/asm/acpi.h | 8 ++ arch/x86/kernel/acpi/boot.c | 10 ++ arch/x86/kernel/mmconf-fam10h_64.c | 2 +- arch/x86/kernel/x86_init.c | 2 +- arch/x86/xen/mmu_pv.c | 5 +- arch/x86/xen/p2m.c | 95 ++++++++++++++ arch/x86/xen/setup.c | 203 ++++++++++++++++++++++------- arch/x86/xen/xen-ops.h | 6 +- 8 files changed, 279 insertions(+), 52 deletions(-)
Signature · 1 line
2.43.0
Juergen Gross
maintainerPATCH v2 1/7 xen: use correct end address of kernel for conflict checking
When running as a Xen PV dom0 the kernel is loaded by the hypervisor using a different memory map than that of the host. In order to minimize the required changes in the kernel, the kernel adapts its memory map to that of the host. In order to do that it is checking for conflicts of its load address with the host memory map.
Unfortunately the tested memory range does not include the .brk area, which might result in crashes or memory corruption when this area does conflict withe the memory map of the host.
Fix the test by using the _end label instead of __bss_stop.
--- arch/x86/xen/setup.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/xen/setup.c +1 -1
@@ -825,7 +825,7 @@ char * __init xen_memory_setup(void)* to relocating (and even reusing) pages with kernel text or data. */ if (xen_is_e820_reserved (__pa_symbol (_text), -__pa_symbol (__bss_stop) -__pa_symbol (_text))) { +__pa_symbol (_end) -__pa_symbol (_text))) {xen_raw_console_write ("Xen hypervisor allocated kernel memory conflicts with E820 map\n" );BUG (); }
Signature · 1 line
2.43.0
Re: [PATCH v2 1/7] xen: use correct end address of kernel for conflict checking
Quoted 10 lines · Juergen Gross
When running as a Xen PV dom0 the kernel is loaded by the hypervisor using a different memory map than that of the host. In order to minimize the required changes in the kernel, the kernel adapts its memory map to that of the host. In order to do that it is checking for conflicts of its load address with the host memory map.
Unfortunately the tested memory range does not include the .brk area, which might result in crashes or memory corruption when this area does conflict withe the memory map of the host.
Quoted 6 lines
Fix the test by using the
_endlabel instead of__bss_stop.Fixes 808fdb71936c ("xen: check for kernel memory conflicting with memory layout")Signed-off-by Juergen Gross <[email protected]>
Juergen Gross
maintainerPATCH v2 2/7 xen: introduce generic helper checking for memory map conflicts
In reply to Juergen GrossWhen booting as a Xen PV dom0 the memory layout of the dom0 is modified to match that of the host, as this requires less changes in the kernel for supporting Xen.
There are some cases, though, which are problematic, as it is the Xen hypervisor selecting the kernel's load address plus some other data, which might conflict with the host's memory map.
These conflicts are detected at boot time and result in a boot error. In order to support handling at least some of these conflicts in future, introduce a generic helper function which will later gain the ability to adapt the memory layout when possible.
Add the missing check for the xen_start_info area.
--- arch/x86/xen/mmu_pv.c | 5 +---- arch/x86/xen/setup.c | 34 ++++++++++++++++++++++++++++------ arch/x86/xen/xen-ops.h | 3 ++- 3 files changed, 31 insertions(+), 11 deletions(-)
arch/x86/xen/mmu_pv.c +1 -4
@@ -2018,10 +2018,7 @@ void __init xen_reserve_special_pages(void)void __init xen_pt_check_e820 (void ) { -if (xen_is_e820_reserved (xen_pt_base, xen_pt_size)) { -xen_raw_console_write ("Xen hypervisor allocated page table memory conflicts with E820 map\n" ); -BUG (); - } +xen_chk_is_e820_usable (xen_pt_base, xen_pt_size,"page table" ); } static unsigned char dummy_mapping[PAGE_SIZE] __page_aligned_bss;
arch/x86/xen/setup.c +28 -6
@@ -567,7 +567,7 @@ static void __init xen_ignore_unusable(void) } } -bool __init xen_is_e820_reserved (phys_addr_t start,phys_addr_t size) +static bool __init xen_is_e820_reserved (phys_addr_t start,phys_addr_t size) {struct e820_entry *entry;unsigned mapcnt;
@@ -624,6 +624,23 @@ phys_addr_t __init xen_find_free_area(phys_addr_t size)return 0 ; } +/* +* Check for an area in physical memory to be usable for non-movable purposes. +* An area is considered to usable if the used E820 map lists it to be RAM. +* In case the area is not usable, crash the system with an error message. +*/ +void __init xen_chk_is_e820_usable (phys_addr_t start,phys_addr_t size, +const char *component) +{ +if (!xen_is_e820_reserved (start, size)) +return ; + +xen_raw_console_write ("Xen hypervisor allocated " ); +xen_raw_console_write (component); +xen_raw_console_write (" memory conflicts with E820 map\n" ); +BUG (); +} +/* * Like memcpy, but with physical addresses for dest and src. */
@@ -824,11 +841,16 @@ char * __init xen_memory_setup(void)* Failing now is better than running into weird problems later due * to relocating (and even reusing) pages with kernel text or data. */ -if (xen_is_e820_reserved (__pa_symbol (_text), -__pa_symbol (_end) -__pa_symbol (_text))) { -xen_raw_console_write ("Xen hypervisor allocated kernel memory conflicts with E820 map\n" ); -BUG (); - } +xen_chk_is_e820_usable (__pa_symbol (_text), +__pa_symbol (_end) -__pa_symbol (_text), +"kernel" ); + +/* +* Check for a conflict of the xen_start_info memory with the target +* E820 map. +*/ +xen_chk_is_e820_usable (__pa (xen_start_info),sizeof (*xen_start_info), +"xen_start_info" );/* * Check for a conflict of the hypervisor supplied page tables with
arch/x86/xen/xen-ops.h +2 -1
@@ -48,7 +48,8 @@ void xen_mm_unpin_all(void);void __init xen_relocate_p2m (void );#endif -bool __init xen_is_e820_reserved (phys_addr_t start,phys_addr_t size); +void __init xen_chk_is_e820_usable (phys_addr_t start,phys_addr_t size, +const char *component);unsigned long __ref xen_chk_extra_mem (unsigned long pfn);void __init xen_inv_extra_mem (void );void __init xen_remap_memory (void );
Signature · 1 line
2.43.0
Re: [PATCH v2 2/7] xen: introduce generic helper checking for memory map conflicts
Quoted 13 lines · Juergen Gross
When booting as a Xen PV dom0 the memory layout of the dom0 is modified to match that of the host, as this requires less changes in the kernel for supporting Xen.
There are some cases, though, which are problematic, as it is the Xen hypervisor selecting the kernel's load address plus some other data, which might conflict with the host's memory map.
These conflicts are detected at boot time and result in a boot error. In order to support handling at least some of these conflicts in future, introduce a generic helper function which will later gain the ability to adapt the memory layout when possible.
Quoted 4 lines
Add the missing check for the
xen_start_infoarea.Signed-off-by Juergen Gross <[email protected]>
However, since you mention the start_info area it may be worth adding half a sentence to the description also mentioning why the hypervisor allocated stack page doesn't need checking. In fact this may want to extend to initrd and phys-mach map as well, to cover everything Xen sets up on behalf of the kernel.
Jan
Jürgen Groß
maintainerRe: [PATCH v2 2/7] xen: introduce generic helper checking for memory map conflicts
Quoted 19 lines · Jan Beulich
Quoted 13 lines · Juergen Gross
When booting as a Xen PV dom0 the memory layout of the dom0 is modified to match that of the host, as this requires less changes in the kernel for supporting Xen.
There are some cases, though, which are problematic, as it is the Xen hypervisor selecting the kernel's load address plus some other data, which might conflict with the host's memory map.
These conflicts are detected at boot time and result in a boot error. In order to support handling at least some of these conflicts in future, introduce a generic helper function which will later gain the ability to adapt the memory layout when possible.
Quoted 4 lines
Add the missing check for the
xen_start_infoarea.Signed-off-by Juergen Gross <[email protected]>
Quoted 7 lines
Reviewed-by Jan Beulich <[email protected]>However, since you mention the
start_infoarea it may be worth adding half a sentence to the description also mentioning why the hypervisor allocated stack page doesn't need checking. In fact this may want to extend to initrd and phys-mach map as well, to cover everything Xen sets up on behalf of the kernel.
Okay.
Juergen
Juergen Gross
maintainerPATCH v2 3/7 xen: move checks for e820 conflicts further up
In reply to Juergen GrossMove the checks for e820 memory map conflicts using the xen_chk_is_e820_usable() helper further up in order to prepare resolving some of the possible conflicts by doing some e820 map modifications, which must happen before evaluating the RAM layout.
--- arch/x86/xen/setup.c | 44 ++++++++++++++++++++++---------------------- 1 file changed, 22 insertions(+), 22 deletions(-)
arch/x86/xen/setup.c +22 -22
@@ -764,6 +764,28 @@ char * __init xen_memory_setup(void)/* Make sure the Xen-supplied memory map is well-ordered. */ e820__update_table (&xen_e820_table); +/* +* Check whether the kernel itself conflicts with the target E820 map. +* Failing now is better than running into weird problems later due +* to relocating (and even reusing) pages with kernel text or data. +*/ +xen_chk_is_e820_usable (__pa_symbol (_text), +__pa_symbol (_end) -__pa_symbol (_text), +"kernel" ); + +/* +* Check for a conflict of the xen_start_info memory with the target +* E820 map. +*/ +xen_chk_is_e820_usable (__pa (xen_start_info),sizeof (*xen_start_info), +"xen_start_info" ); + +/* +* Check for a conflict of the hypervisor supplied page tables with +* the target E820 map. +*/ +xen_pt_check_e820 (); + max_pages =xen_get_max_pages ();/* How many extra pages do we need due to remapping? */
@@ -836,28 +858,6 @@ char * __init xen_memory_setup(void)e820__update_table (e820_table); -/* -* Check whether the kernel itself conflicts with the target E820 map. -* Failing now is better than running into weird problems later due -* to relocating (and even reusing) pages with kernel text or data. -*/ -xen_chk_is_e820_usable (__pa_symbol (_text), -__pa_symbol (_end) -__pa_symbol (_text), -"kernel" ); - -/* -* Check for a conflict of the xen_start_info memory with the target -* E820 map. -*/ -xen_chk_is_e820_usable (__pa (xen_start_info),sizeof (*xen_start_info), -"xen_start_info" ); - -/* -* Check for a conflict of the hypervisor supplied page tables with -* the target E820 map. -*/ -xen_pt_check_e820 (); -xen_reserve_xen_mfnlist ();/* Check for a conflict of the initrd with the target E820 map. */
Signature · 1 line
2.43.0
Re: [PATCH v2 3/7] xen: move checks for e820 conflicts further up
Quoted 7 lines · Juergen Gross
Move the checks for e820 memory map conflicts using the
xen_chk_is_e820_usable()helper further up in order to prepare resolving some of the possible conflicts by doing some e820 map modifications, which must happen before evaluating the RAM layout.Signed-off-by Juergen Gross <[email protected]>
Juergen Gross
maintainerPATCH v2 4/7 xen: move max_pfn in xen_memory_setup() out of function scope
In reply to Juergen GrossInstead of having max_pfn as a local variable of xen_memory_setup(), make it a static variable in setup.c instead. This avoids having to pass it to subfunctions, which will be needed in more cases in future.
Rename it to ini_nr_pages, as the value denotes the currently usable number of memory pages as passed from the hypervisor at boot time.
--- arch/x86/xen/setup.c | 53 ++++++++++++++++++++++---------------------- 1 file changed, 27 insertions(+), 26 deletions(-)
arch/x86/xen/setup.c +27 -26
@@ -46,6 +46,9 @@ bool xen_pv_pci_possible;/* E820 map used during setting up memory. */ static struct e820_table xen_e820_table__initdata ; +/* Number of initially usable memory pages. */ +static unsigned long ini_nr_pages__initdata ; +/* * Buffer used to remap identity mapped pages. We only need the virtual space. * The physical page behind this address is remapped as needed to different
@@ -212,7 +215,7 @@ static int __init xen_free_mfn(unsigned long mfn)* as a fallback if the remapping fails. */ static void __init xen_set_identity_and_release_chunk (unsigned long start_pfn, -unsigned long end_pfn,unsigned long nr_pages) +unsigned long end_pfn) {unsigned long pfn, end;int ret;
@@ -220,7 +223,7 @@ static void __init xen_set_identity_and_release_chunk(unsigned long start_pfn,WARN_ON (start_pfn > end_pfn);/* Release pages first. */ - end =min (end_pfn, nr_pages); + end =min (end_pfn, ini_nr_pages);for (pfn = start_pfn; pfn < end; pfn++) {unsigned long mfn =pfn_to_mfn (pfn);
@@ -341,15 +344,14 @@ static void __init xen_do_set_identity_and_remap_chunk(* to Xen and not remapped. */ static unsigned long __init xen_set_identity_and_remap_chunk ( -unsigned long start_pfn,unsigned long end_pfn,unsigned long nr_pages, -unsigned long remap_pfn) +unsigned long start_pfn,unsigned long end_pfn,unsigned long remap_pfn) {unsigned long pfn;unsigned long i =0 ;unsigned long n = end_pfn - start_pfn;if (remap_pfn ==0 ) - remap_pfn = nr_pages; + remap_pfn = ini_nr_pages;while (i < n) {unsigned long cur_pfn = start_pfn + i;
@@ -358,19 +360,19 @@ static unsigned long __init xen_set_identity_and_remap_chunk(unsigned long remap_range_size;/* Do not remap pages beyond the current allocation */ -if (cur_pfn >= nr_pages) { +if (cur_pfn >= ini_nr_pages) {/* Identity map remaining pages */ set_phys_range_identity (cur_pfn, cur_pfn + size);break ; } -if (cur_pfn + size > nr_pages) - size = nr_pages - cur_pfn; +if (cur_pfn + size > ini_nr_pages) + size = ini_nr_pages - cur_pfn; remap_range_size =xen_find_pfn_range (&remap_pfn);if (!remap_range_size) {pr_warn ("Unable to find available pfn range, not remapping identity pages\n" );xen_set_identity_and_release_chunk (cur_pfn, - cur_pfn + left, nr_pages); + cur_pfn + left);break ; }/* Adjust size to fit in current e820 RAM region */
@@ -397,18 +399,18 @@ static unsigned long __init xen_set_identity_and_remap_chunk( }static unsigned long __init xen_count_remap_pages ( -unsigned long start_pfn,unsigned long end_pfn,unsigned long nr_pages, +unsigned long start_pfn,unsigned long end_pfn,unsigned long remap_pages) { -if (start_pfn >= nr_pages) +if (start_pfn >= ini_nr_pages)return remap_pages; -return remap_pages +min (end_pfn, nr_pages) - start_pfn; +return remap_pages +min (end_pfn, ini_nr_pages) - start_pfn; } -static unsigned long __init xen_foreach_remap_area (unsigned long nr_pages, +static unsigned long __init xen_foreach_remap_area (unsigned long (*func)(unsigned long start_pfn,unsigned long end_pfn, -unsigned long nr_pages,unsigned long last_val)) +unsigned long last_val)) {phys_addr_t start =0 ;unsigned long ret_val =0 ;
@@ -436,8 +438,7 @@ static unsigned long __init xen_foreach_remap_area(unsigned long nr_pages, end_pfn =PFN_UP (entry->addr);if (start_pfn < end_pfn) - ret_val =func (start_pfn, end_pfn, nr_pages, - ret_val); + ret_val =func (start_pfn, end_pfn, ret_val); start = end; } }
@@ -700,7 +701,7 @@ static void __init xen_reserve_xen_mfnlist(void)**/ char *__init xen_memory_setup (void ) { -unsigned long max_pfn, pfn_s, n_pfns; +unsigned long pfn_s, n_pfns;phys_addr_t mem_end, addr, size, chunk_size;u32 type;int rc;
@@ -712,9 +713,9 @@ char * __init xen_memory_setup(void)int op;xen_parse_512gb (); - max_pfn =xen_get_pages_limit (); - max_pfn =min (max_pfn, xen_start_info->nr_pages); - mem_end =PFN_PHYS (max_pfn); + ini_nr_pages =xen_get_pages_limit (); + ini_nr_pages =min (ini_nr_pages, xen_start_info->nr_pages); + mem_end =PFN_PHYS (ini_nr_pages); memmap.nr_entries =ARRAY_SIZE (xen_e820_table.entries);set_xen_guest_handle (memmap.buffer, xen_e820_table.entries);
@@ -789,10 +790,10 @@ char * __init xen_memory_setup(void) max_pages =xen_get_max_pages ();/* How many extra pages do we need due to remapping? */ - max_pages +=xen_foreach_remap_area (max_pfn, xen_count_remap_pages); + max_pages +=xen_foreach_remap_area (xen_count_remap_pages); -if (max_pages > max_pfn) - extra_pages += max_pages - max_pfn; +if (max_pages > ini_nr_pages) + extra_pages += max_pages - ini_nr_pages;/* * Clamp the amount of extra memory to a EXTRA_MEM_RATIO
@@ -801,8 +802,8 @@ char * __init xen_memory_setup(void)* Make sure we have no memory above max_pages, as this area * isn't handled by the p2m management. */ - maxmem_pages =EXTRA_MEM_RATIO *min (max_pfn,PFN_DOWN (MAXMEM )); - extra_pages =min3 (maxmem_pages, extra_pages, max_pages - max_pfn); + maxmem_pages =EXTRA_MEM_RATIO *min (ini_nr_pages,PFN_DOWN (MAXMEM )); + extra_pages =min3 (maxmem_pages, extra_pages, max_pages - ini_nr_pages); i =0 ; addr = xen_e820_table.entries[0 ].addr; size = xen_e820_table.entries[0 ].size;
@@ -885,7 +886,7 @@ char * __init xen_memory_setup(void)* Set identity map on non-RAM pages and prepare remapping the * underlying RAM. */ -xen_foreach_remap_area (max_pfn, xen_set_identity_and_remap_chunk); +xen_foreach_remap_area (xen_set_identity_and_remap_chunk);pr_info ("Released %ld page(s)\n" , xen_released_pages);
Signature · 1 line
2.43.0
Re: [PATCH v2 4/7] xen: move max_pfn in xen_memory_setup() out of function scope
Quoted 9 lines · Juergen Gross
Instead of having
max_pfnas a local variable ofxen_memory_setup(), make it a static variable insetup.cinstead. This avoids having to pass it to subfunctions, which will be needed in more cases in future.Rename it to
ini_nr_pages, as the value denotes the currently usable number of memory pages as passed from the hypervisor at boot time.Signed-off-by Juergen Gross <[email protected]>
I wonder though why ...
arch/x86/xen/setup.c char * __init xen_memory_setup(void) · line 717
Quoted 9 lines
arch/x86/xen/setup.c +2
-3@@ -712,9 +713,9 @@ char * __init xen_memory_setup(void)int op;xen_parse_512gb (); - max_pfn =xen_get_pages_limit (); - max_pfn =min (max_pfn, xen_start_info->nr_pages); - mem_end =PFN_PHYS (max_pfn); + ini_nr_pages =xen_get_pages_limit (); + ini_nr_pages =min (ini_nr_pages, xen_start_info->nr_pages);
... this can't be just
ini_nr_pages = min(xen_get_pages_limit(), xen_start_info->nr_pages);
Jan
Jürgen Groß
maintainerRe: [PATCH v2 4/7] xen: move max_pfn in xen_memory_setup() out of function scope
Quoted 27 lines · Jan Beulich
Quoted 9 lines · Juergen Gross
Instead of having
max_pfnas a local variable ofxen_memory_setup(), make it a static variable insetup.cinstead. This avoids having to pass it to subfunctions, which will be needed in more cases in future.Rename it to
ini_nr_pages, as the value denotes the currently usable number of memory pages as passed from the hypervisor at boot time.Signed-off-by Juergen Gross <[email protected]>Reviewed-by Jan Beulich <[email protected]>I wonder though why ...
arch/x86/xen/setup.c char * __init xen_memory_setup(void) · line 717
Quoted 9 lines
arch/x86/xen/setup.c +2
-3@@ -712,9 +713,9 @@ char * __init xen_memory_setup(void)int op;xen_parse_512gb (); - max_pfn =xen_get_pages_limit (); - max_pfn =min (max_pfn, xen_start_info->nr_pages); - mem_end =PFN_PHYS (max_pfn); + ini_nr_pages =xen_get_pages_limit (); + ini_nr_pages =min (ini_nr_pages, xen_start_info->nr_pages);... this can't be just
Quoted 1 line
ini_nr_pages = min(xen_get_pages_limit(), xen_start_info->nr_pages);
I just did a mechanical change for all uses of max_pfn.
I agree that your suggestion is an improvement, so I'll modify the code.
Juergen
Juergen Gross
maintainerPATCH v2 5/7 xen: add capability to remap non-RAM pages to different PFNs
In reply to Juergen GrossWhen running as a Xen PV dom0 it can happen that the kernel is being loaded to a guest physical address conflicting with the host memory map.
In order to be able to resolve this conflict, add the capability to remap non-RAM areas to different guest PFNs. A function to use this remapping information for other purposes than doing the remap will be added when needed.
As the number of conflicts should be rather low (currently only machines with max. 1 conflict are known), save the remap data in a small statically allocated array.
---
V2:
- split off from patch 5 of V1 of the series
- moved to p2m.c
--- arch/x86/xen/p2m.c | 65 ++++++++++++++++++++++++++++++++++++++++++ arch/x86/xen/xen-ops.h | 3 ++ 2 files changed, 68 insertions(+)
arch/x86/xen/p2m.c +65 -0
@@ -80,6 +80,7 @@#include <asm/xen/hypervisor.h> #include <xen/balloon.h> #include <xen/grant_table.h> +#include <xen/hvc-console.h> #include "xen-ops.h"
@@ -792,6 +793,70 @@ int clear_foreign_p2m_mapping(struct gnttab_unmap_grant_ref *unmap_ops,return ret; } +/* Remapped non-RAM areas */ +#define NR_NONRAM_REMAP 4 +static struct nonram_remap { +phys_addr_t maddr; +phys_addr_t paddr; +unsigned long size; +} xen_nonram_remap[NR_NONRAM_REMAP ]; +static unsigned int nr_nonram_remap; + +/* +* Do the real remapping of non-RAM regions as specified in the +* xen_nonram_remap[] array. +* In case of an error just crash the system. +*/ +void __init xen_do_remap_nonram (void ) +{ +unsigned int i; +unsigned int remapped =0 ; +struct nonram_remap *remap = xen_nonram_remap; +unsigned long pfn, mfn, len; + +if (!nr_nonram_remap) +return ; + +for (i =0 ; i < nr_nonram_remap; i++) { + pfn =PFN_DOWN (remap->paddr); + mfn =PFN_DOWN (remap->maddr); +for (len =0 ; len < remap->size; len +=PAGE_SIZE ) { +if (!set_phys_to_machine (pfn, mfn)) { +pr_err ("Failed to set p2m mapping for pfn=%ld mfn=%ld\n" , + pfn, mfn); +BUG (); + } + + pfn++; + mfn++; + remapped++; + } + + remap++; + } + +pr_info ("Remapped %u non-RAM page(s)\n" , remapped); +} + +/* +* Add a new non-RAM remap entry. +* In case of no free entry found, just crash the system. +*/ +void __init xen_add_remap_nonram (phys_addr_t maddr,phys_addr_t paddr, +unsigned long size) +{ +if (nr_nonram_remap ==NR_NONRAM_REMAP ) { +xen_raw_console_write ("Number of required E820 entry remapping actions exceed maximum value\n" ); +BUG (); + } + + xen_nonram_remap[nr_nonram_remap].maddr = maddr; + xen_nonram_remap[nr_nonram_remap].paddr = paddr; + xen_nonram_remap[nr_nonram_remap].size = size; + + nr_nonram_remap++; +} +#ifdef CONFIG_XEN_DEBUG_FS #include <linux/debugfs.h> static int p2m_dump_show (struct seq_file *m,void *v)
arch/x86/xen/xen-ops.h +3 -0
@@ -47,6 +47,9 @@ void xen_mm_unpin_all(void);#ifdef CONFIG_X86_64 void __init xen_relocate_p2m (void );#endif +void __init xen_do_remap_nonram (void ); +void __init xen_add_remap_nonram (phys_addr_t maddr,phys_addr_t paddr, +unsigned long size);void __init xen_chk_is_e820_usable (phys_addr_t start,phys_addr_t size,const char *component);
Signature · 1 line
2.43.0
Re: [PATCH v2 5/7] xen: add capability to remap non-RAM pages to different PFNs
Quoted 36 lines · Juergen Gross
When running as a Xen PV dom0 it can happen that the kernel is being loaded to a guest physical address conflicting with the host memory map.
In order to be able to resolve this conflict, add the capability to remap non-RAM areas to different guest PFNs. A function to use this remapping information for other purposes than doing the remap will be added when needed.
As the number of conflicts should be rather low (currently only machines with max. 1 conflict are known), save the remap data in a small statically allocated array.
Signed-off-by Juergen Gross <[email protected]>---
V2:
- split off from patch 5 of V1 of the series
- moved top2m.c--- arch/x86/xen/p2m.c | 65 ++++++++++++++++++++++++++++++++++++++++++ arch/x86/xen/xen-ops.h | 3 ++ 2 files changed, 68 insertions(+)arch/x86/xen/p2m.c +1
-0@@ -80,6 +80,7 @@#include <asm/xen/hypervisor.h> #include <xen/balloon.h> #include <xen/grant_table.h> +#include <xen/hvc-console.h> #include "xen-ops.h"
arch/x86/xen/p2m.c int clear_foreign_p2m_mapping(struct gnttab_unmap_grant_ref *unmap_ops, · line 801
Quoted 9 lines
arch/x86/xen/p2m.c +6
-0@@ -792,6 +793,70 @@ int clear_foreign_p2m_mapping(struct gnttab_unmap_grant_ref *unmap_ops,return ret; } +/* Remapped non-RAM areas */ +#define NR_NONRAM_REMAP 4 +static struct nonram_remap { + phys_addr_t maddr; + phys_addr_t paddr; + unsigned long size;
size_t?
arch/x86/xen/p2m.c int clear_foreign_p2m_mapping(struct gnttab_unmap_grant_ref *unmap_ops, · line 803
Quoted 2 lines
arch/x86/xen/p2m.c +2
-0+} xen_nonram_remap[NR_NONRAM_REMAP ]; +static unsigned int nr_nonram_remap;
Both __initdata? Or, considering patch 6, at least __ro_after_init?
Quoted 11 lines
arch/x86/xen/p2m.c +11
-0+/* +* Do the real remapping of non-RAM regions as specified in the +* xen_nonram_remap[] array. +* In case of an error just crash the system. +*/ +void __init xen_do_remap_nonram (void ) +{ +unsigned int i; +unsigned int remapped =0 ; +struct nonram_remap *remap = xen_nonram_remap; +unsigned long pfn, mfn, len;
arch/x86/xen/p2m.c int clear_foreign_p2m_mapping(struct gnttab_unmap_grant_ref *unmap_ops, · line 825
Quoted 10 lines
arch/x86/xen/p2m.c +10
-0+ +if (!nr_nonram_remap) +return ; + +for (i =0 ; i < nr_nonram_remap; i++) { + pfn =PFN_DOWN (remap->paddr); + mfn =PFN_DOWN (remap->maddr); +for (len =0 ; len < remap->size; len +=PAGE_SIZE ) { +if (!set_phys_to_machine (pfn, mfn)) { + pr_err("Failed to set p2m mapping for pfn=%ld mfn=%ld\n",
I'm not convinced that frame numbers logged in decimal are overly useful.
arch/x86/xen/p2m.c int clear_foreign_p2m_mapping(struct gnttab_unmap_grant_ref *unmap_ops, · line 838
Quoted 13 lines
arch/x86/xen/p2m.c +13
-0+ pfn, mfn); +BUG (); + } + + pfn++; + mfn++; + remapped++; + } + + remap++; + } + +pr_info ("Remapped %u non-RAM page(s)\n" , remapped);
This message may be useful in a log also when nothing was remapped - maybe drop the initial if()?
Quoted 10 lines
arch/x86/xen/p2m.c +10
-0+} + +/* +* Add a new non-RAM remap entry. +* In case of no free entry found, just crash the system. +*/ +void __init xen_add_remap_nonram (phys_addr_t maddr,phys_addr_t paddr, +unsigned long size) +{ +if (nr_nonram_remap ==NR_NONRAM_REMAP ) {
arch/x86/xen/p2m.c int clear_foreign_p2m_mapping(struct gnttab_unmap_grant_ref *unmap_ops, · line 858
Quoted 10 lines
arch/x86/xen/p2m.c +10
-0+xen_raw_console_write ("Number of required E820 entry remapping actions exceed maximum value\n" ); +BUG (); + } + + xen_nonram_remap[nr_nonram_remap].maddr = maddr; + xen_nonram_remap[nr_nonram_remap].paddr = paddr; + xen_nonram_remap[nr_nonram_remap].size = size; + + nr_nonram_remap++; +}
You don't enforce any constraints on the addresses / size here. With this the loop in xen_do_remap_nonram() may terminate too early if non- page-aligned values were passed into here. Both addresses not having the same offset-into-page may also end up anomalous. Might be worth switching to frame numbers / number-of-pages for the tracking struct.
Jan
Jürgen Groß
maintainerRe: [PATCH v2 5/7] xen: add capability to remap non-RAM pages to different PFNs
Quoted 40 lines · Jan Beulich
Quoted 39 lines · Juergen Gross
When running as a Xen PV dom0 it can happen that the kernel is being loaded to a guest physical address conflicting with the host memory map.
In order to be able to resolve this conflict, add the capability to remap non-RAM areas to different guest PFNs. A function to use this remapping information for other purposes than doing the remap will be added when needed.
As the number of conflicts should be rather low (currently only machines with max. 1 conflict are known), save the remap data in a small statically allocated array.
Signed-off-by Juergen Gross <[email protected]>---
V2:
- split off from patch 5 of V1 of the series
- moved top2m.c--- arch/x86/xen/p2m.c | 65 ++++++++++++++++++++++++++++++++++++++++++ arch/x86/xen/xen-ops.h | 3 ++ 2 files changed, 68 insertions(+)arch/x86/xen/p2m.c +1
-0@@ -80,6 +80,7 @@#include <asm/xen/hypervisor.h> #include <xen/balloon.h> #include <xen/grant_table.h> +#include <xen/hvc-console.h> #include "xen-ops.h" @@ -792,6 +793,70 @@ int clear_foreign_p2m_mapping(struct gnttab_unmap_grant_ref *unmap_ops,return ret; }
Quoted 8 lines
arch/x86/xen/p2m.c int clear_foreign_p2m_mapping(struct gnttab_unmap_grant_ref *unmap_ops, · line 801
Quoted 6 lines
arch/x86/xen/p2m.c +6
-0+/* Remapped non-RAM areas */ +#define NR_NONRAM_REMAP 4 +static struct nonram_remap { + phys_addr_t maddr; + phys_addr_t paddr; + unsigned long size;
size_t?
Fine with me.
Quoted 5 lines
arch/x86/xen/p2m.c int clear_foreign_p2m_mapping(struct gnttab_unmap_grant_ref *unmap_ops, · line 803
Quoted 3 lines
arch/x86/xen/p2m.c +2
-0+} xen_nonram_remap[NR_NONRAM_REMAP ]; +static unsigned int nr_nonram_remap;Both
__initdata? Or, considering patch 6, at least__ro_after_init?
__ro_after_init should be fine.
Quoted 16 lines
Quoted 16 lines
arch/x86/xen/p2m.c +15
-0+/* +* Do the real remapping of non-RAM regions as specified in the +* xen_nonram_remap[] array. +* In case of an error just crash the system. +*/ +void __init xen_do_remap_nonram (void ) +{ +unsigned int i; +unsigned int remapped =0 ; +struct nonram_remap *remap = xen_nonram_remap; +unsigned long pfn, mfn, len; + +if (!nr_nonram_remap) +return ; +
Quoted 8 lines
arch/x86/xen/p2m.c int clear_foreign_p2m_mapping(struct gnttab_unmap_grant_ref *unmap_ops, · line 825
Quoted 6 lines
arch/x86/xen/p2m.c +6
-0+for (i =0 ; i < nr_nonram_remap; i++) { + pfn =PFN_DOWN (remap->paddr); + mfn =PFN_DOWN (remap->maddr); +for (len =0 ; len < remap->size; len +=PAGE_SIZE ) { +if (!set_phys_to_machine (pfn, mfn)) { + pr_err("Failed to set p2m mapping for pfn=%ld mfn=%ld\n",I'm not convinced that frame numbers logged in decimal are overly useful.
I agree. Will switch to hex.
Quoted 17 lines
Quoted 4 lines
arch/x86/xen/p2m.c +3
-0+ pfn, mfn); +BUG (); + }arch/x86/xen/p2m.c int clear_foreign_p2m_mapping(struct gnttab_unmap_grant_ref *unmap_ops, · line 838
Quoted 10 lines
arch/x86/xen/p2m.c +10
-0+ + pfn++; + mfn++; + remapped++; + } + + remap++; + } + +pr_info ("Remapped %u non-RAM page(s)\n" , remapped);This message may be useful in a log also when nothing was remapped - maybe drop the initial
if()?
Fine with me.
Quoted 19 lines
Quoted 19 lines
arch/x86/xen/p2m.c +18
-0+} + +/* +* Add a new non-RAM remap entry. +* In case of no free entry found, just crash the system. +*/ +void __init xen_add_remap_nonram (phys_addr_t maddr,phys_addr_t paddr, +unsigned long size) +{ +if (nr_nonram_remap ==NR_NONRAM_REMAP ) { +xen_raw_console_write ("Number of required E820 entry remapping actions exceed maximum value\n" ); +BUG (); + } + + xen_nonram_remap[nr_nonram_remap].maddr = maddr; + xen_nonram_remap[nr_nonram_remap].paddr = paddr; + xen_nonram_remap[nr_nonram_remap].size = size; +
Quoted 8 lines
arch/x86/xen/p2m.c int clear_foreign_p2m_mapping(struct gnttab_unmap_grant_ref *unmap_ops, · line 858
Quoted 2 lines
arch/x86/xen/p2m.c +2
-0+ nr_nonram_remap++; +}You don't enforce any constraints on the addresses / size here. With this the loop in
xen_do_remap_nonram()may terminate too early if non- page-aligned values were passed into here. Both addresses not having the same offset-into-page may also end up anomalous. Might be worth switching to frame numbers / number-of-pages for the tracking struct.
Hmm, I'd like to at least WARN() in case someone tries to access a remapped area out of bounds, which requires to keep the original non-page-aligned values.
I'll fix the xen_do_remap_nonram() loop.
Juergen
Juergen Gross
maintainerPATCH v2 6/7 xen: allow mapping ACPI data using a different physical address
In reply to Juergen GrossWhen running as a Xen PV dom0 the system needs to map ACPI data of the host using host physical addresses, while those addresses can conflict with the guest physical addresses of the loaded linux kernel.
This conflict can be solved by mapping the ACPI data to a different guest physical address, but mapping the data via acpi_os_ioremap() must still be possible using the host physical address, as this address might be generated by AML when referencing some of the ACPI data.
When configured to support running as a Xen PV dom0, have an implementation of acpi_os_ioremap() being aware of the possibility to need above mentioned translation of a host physical address to the guest physical address.
This modification requires to fix some #include of asm/acpi.h in x86 code to use linux/acpi.h instead.
---
V2:
- new patch (Jan Beulich)
--- arch/x86/include/asm/acpi.h | 8 ++++++++ arch/x86/kernel/acpi/boot.c | 10 ++++++++++ arch/x86/kernel/mmconf-fam10h_64.c | 2 +- arch/x86/kernel/x86_init.c | 2 +- arch/x86/xen/p2m.c | 30 ++++++++++++++++++++++++++++++ arch/x86/xen/setup.c | 2 +- 6 files changed, 51 insertions(+), 3 deletions(-)
arch/x86/include/asm/acpi.h +8 -0
@@ -174,6 +174,14 @@ void acpi_generic_reduced_hw_init(void);void x86_default_set_root_pointer (u64 addr);u64 x86_default_get_root_pointer (void ); +#ifdef CONFIG_XEN_PV_DOM0 +/* A Xen PV dom0 needs a special acpi_os_ioremap() handling. */ +extern void __iomem * (*acpi_os_ioremap)(acpi_physical_address phys, +acpi_size size); +void __iomem *x86_acpi_os_ioremap (acpi_physical_address phys,acpi_size size); +#define acpi_os_ioremap acpi_os_ioremap +#endif +#else /* !CONFIG_ACPI */ #define acpi_lapic 0
arch/x86/kernel/acpi/boot.c +10 -0
@@ -1778,3 +1778,13 @@ u64 x86_default_get_root_pointer(void) {return boot_params.acpi_rsdp_addr; } + +#ifdef CONFIG_XEN_PV_DOM0 +void __iomem *x86_acpi_os_ioremap (acpi_physical_address phys,acpi_size size) +{ +return ioremap_cache (phys, size); +} + +void __iomem * (*acpi_os_ioremap)(acpi_physical_address phys,acpi_size size) = + x86_acpi_os_ioremap; +#endif
arch/x86/kernel/mmconf-fam10h_64.c +1 -1
@@ -9,12 +9,12 @@#include <linux/pci.h> #include <linux/dmi.h> #include <linux/range.h> +#include <linux/acpi.h> #include <asm/pci-direct.h> #include <linux/sort.h> #include <asm/io.h> #include <asm/msr.h> -#include <asm/acpi.h> #include <asm/mmconfig.h> #include <asm/pci_x86.h>
arch/x86/kernel/x86_init.c +1 -1
@@ -8,8 +8,8 @@#include <linux/ioport.h> #include <linux/export.h> #include <linux/pci.h> +#include <linux/acpi.h> -#include <asm/acpi.h> #include <asm/bios_ebda.h> #include <asm/paravirt.h> #include <asm/pci_x86.h>
arch/x86/xen/p2m.c +30 -0
@@ -70,6 +70,7 @@#include <linux/memblock.h> #include <linux/slab.h> #include <linux/vmalloc.h> +#include <linux/acpi.h> #include <asm/cache.h> #include <asm/setup.h>
@@ -838,6 +839,31 @@ void __init xen_do_remap_nonram(void)pr_info ("Remapped %u non-RAM page(s)\n" , remapped); } +/* +* Xen variant of acpi_os_ioremap() taking potentially remapped non-RAM +* regions into acount. +* Any attempt to map an area crossing a remap boundary will produce a +* WARN() splat. +*/ +static void __iomem *xen_acpi_os_ioremap (acpi_physical_address phys, +acpi_size size) +{ +unsigned int i; +struct nonram_remap *remap = xen_nonram_remap; + +for (i =0 ; i < nr_nonram_remap; i++) { +if (phys + size > remap->maddr && + phys < remap->maddr + remap->size) { +WARN_ON (phys < remap->maddr || + phys + size > remap->maddr + remap->size); + phys = remap->paddr + phys - remap->maddr; +break ; + } + } + +return x86_acpi_os_ioremap (phys, size); +} +/* * Add a new non-RAM remap entry. * In case of no free entry found, just crash the system.
@@ -850,6 +876,10 @@ void __init xen_add_remap_nonram(phys_addr_t maddr, phys_addr_t paddr,BUG (); } +/* Switch to the Xen acpi_os_ioremap() variant. */ +if (nr_nonram_remap ==0 ) + acpi_os_ioremap = xen_acpi_os_ioremap; + xen_nonram_remap[nr_nonram_remap].maddr = maddr; xen_nonram_remap[nr_nonram_remap].paddr = paddr; xen_nonram_remap[nr_nonram_remap].size = size;
arch/x86/xen/setup.c +1 -1
@@ -15,12 +15,12 @@#include <linux/cpuidle.h> #include <linux/cpufreq.h> #include <linux/memory_hotplug.h> +#include <linux/acpi.h> #include <asm/elf.h> #include <asm/vdso.h> #include <asm/e820/api.h> #include <asm/setup.h> -#include <asm/acpi.h> #include <asm/numa.h> #include <asm/idtentry.h> #include <asm/xen/hypervisor.h>
Signature · 1 line
2.43.0
Re: [PATCH v2 6/7] xen: allow mapping ACPI data using a different physical address
Quoted 5 lines · Juergen Gross
arch/x86/xen/p2m.c +1
-0@@ -838,6 +839,31 @@ void __init xen_do_remap_nonram(void)pr_info ("Remapped %u non-RAM page(s)\n" , remapped); } +/*
arch/x86/xen/p2m.c void __init xen_do_remap_nonram(void) · line 852
Quoted 10 lines
arch/x86/xen/p2m.c +10
-0+* Xen variant of acpi_os_ioremap() taking potentially remapped non-RAM +* regions into acount. +* Any attempt to map an area crossing a remap boundary will produce a +* WARN() splat. +*/ +static void __iomem *xen_acpi_os_ioremap (acpi_physical_address phys, +acpi_size size) +{ +unsigned int i; +struct nonram_remap *remap = xen_nonram_remap;
const (also in one of the functions in patch 5)?
arch/x86/xen/p2m.c void __init xen_do_remap_nonram(void) · line 865
Quoted 12 lines
arch/x86/xen/p2m.c +12
-0+for (i =0 ; i < nr_nonram_remap; i++) { +if (phys + size > remap->maddr && + phys < remap->maddr + remap->size) { +WARN_ON (phys < remap->maddr || + phys + size > remap->maddr + remap->size); + phys = remap->paddr + phys - remap->maddr; +break ; + } + } + +return x86_acpi_os_ioremap (phys, size); +}
At least this, perhaps also what patch 5 adds, likely wants to be limited to the XEN_DOM0 case? Or else I wonder whether ...
arch/x86/xen/p2m.c void __init xen_add_remap_nonram(phys_addr_t maddr, phys_addr_t paddr, · line 881
Quoted 7 lines
arch/x86/xen/p2m.c +3
-0@@ -850,6 +876,10 @@ void __init xen_add_remap_nonram(phys_addr_t maddr, phys_addr_t paddr,BUG (); } +/* Switch to the Xen acpi_os_ioremap() variant. */ +if (nr_nonram_remap ==0 ) + acpi_os_ioremap = xen_acpi_os_ioremap;
... this would actually build when XEN_DOM0=n.
I'm actually surprised there's no Dom0-only code section in this file, where the new code could then simply be inserted.
Jan
Juergen Gross
maintainerRe: [PATCH v2 6/7] xen: allow mapping ACPI data using a different physical address
Quoted 5 lines · Jan Beulich
Quoted 4 lines · Juergen Gross
arch/x86/xen/p2m.c +0
-0@@ -838,6 +839,31 @@ void __init xen_do_remap_nonram(void)pr_info ("Remapped %u non-RAM page(s)\n" , remapped); }
Quoted 13 lines
arch/x86/xen/p2m.c void __init xen_do_remap_nonram(void) · line 852
Quoted 11 lines
arch/x86/xen/p2m.c +11
-0+/* +* Xen variant of acpi_os_ioremap() taking potentially remapped non-RAM +* regions into acount. +* Any attempt to map an area crossing a remap boundary will produce a +* WARN() splat. +*/ +static void __iomem *xen_acpi_os_ioremap (acpi_physical_address phys, +acpi_size size) +{ +unsigned int i; +struct nonram_remap *remap = xen_nonram_remap;const (also in one of the functions in patch 5)?
Yes.
Quoted 21 lines
arch/x86/xen/p2m.c void __init xen_do_remap_nonram(void) · line 865
Quoted 13 lines
arch/x86/xen/p2m.c +12
-0+for (i =0 ; i < nr_nonram_remap; i++) { +if (phys + size > remap->maddr && + phys < remap->maddr + remap->size) { +WARN_ON (phys < remap->maddr || + phys + size > remap->maddr + remap->size); + phys = remap->paddr + phys - remap->maddr; +break ; + } + } + +return x86_acpi_os_ioremap (phys, size); +}At least this, perhaps also what patch 5 adds, likely wants to be limited to the
XEN_DOM0case? Or else I wonder whether ...Quoted 4 lines
(quoted diff) +0
-0@@ -850,6 +876,10 @@ void __init xen_add_remap_nonram(phys_addr_t maddr, phys_addr_t paddr, BUG(); }
Quoted 8 lines
arch/x86/xen/p2m.c void __init xen_add_remap_nonram(phys_addr_t maddr, phys_addr_t paddr, · line 881
Quoted 3 lines
arch/x86/xen/p2m.c +3
-0+/* Switch to the Xen acpi_os_ioremap() variant. */ +if (nr_nonram_remap ==0 ) + acpi_os_ioremap = xen_acpi_os_ioremap;... this would actually build when XEN_DOM0=n.
I'm actually surprised there's no Dom0-only code section in this file, where the new code could then simply be inserted.
I'd rather make this conditional on CONFIG_ACPI.
Depending on how Xen tools will handle a PV-domain with "e820_host=1" this code might be important for domUs, too.
Juergen
Re: [PATCH v2 6/7] xen: allow mapping ACPI data using a different physical address
Quoted 49 lines · Juergen Gross
Quoted 5 lines · Jan Beulich
Quoted 4 lines · Juergen Gross
arch/x86/xen/p2m.c +0
-0@@ -838,6 +839,31 @@ void __init xen_do_remap_nonram(void)pr_info ("Remapped %u non-RAM page(s)\n" , remapped); }Quoted 13 lines
arch/x86/xen/p2m.c void __init xen_do_remap_nonram(void) · line 852
Quoted 11 lines
arch/x86/xen/p2m.c +11
-0+/* +* Xen variant of acpi_os_ioremap() taking potentially remapped non-RAM +* regions into acount. +* Any attempt to map an area crossing a remap boundary will produce a +* WARN() splat. +*/ +static void __iomem *xen_acpi_os_ioremap (acpi_physical_address phys, +acpi_size size) +{ +unsigned int i; +struct nonram_remap *remap = xen_nonram_remap;const (also in one of the functions in patch 5)?
Yes.
Quoted 27 lines
arch/x86/xen/p2m.c void __init xen_do_remap_nonram(void) · line 865
Quoted 13 lines
arch/x86/xen/p2m.c +12
-0+for (i =0 ; i < nr_nonram_remap; i++) { +if (phys + size > remap->maddr && + phys < remap->maddr + remap->size) { +WARN_ON (phys < remap->maddr || + phys + size > remap->maddr + remap->size); + phys = remap->paddr + phys - remap->maddr; +break ; + } + } + +return x86_acpi_os_ioremap (phys, size); +}At least this, perhaps also what patch 5 adds, likely wants to be limited to the
XEN_DOM0case? Or else I wonder whether ...arch/x86/xen/p2m.c void __init xen_add_remap_nonram(phys_addr_t maddr, phys_addr_t paddr, · line 881
Quoted 7 lines
arch/x86/xen/p2m.c +3
-0@@ -850,6 +876,10 @@ void __init xen_add_remap_nonram(phys_addr_t maddr, phys_addr_t paddr,BUG (); } +/* Switch to the Xen acpi_os_ioremap() variant. */ +if (nr_nonram_remap ==0 ) + acpi_os_ioremap = xen_acpi_os_ioremap;... this would actually build when XEN_DOM0=n.
Quoted 4 lines
Quoted 2 lines
I'm actually surprised there's no Dom0-only code section in this file, where the new code could then simply be inserted.
I'd rather make this conditional on
CONFIG_ACPI.
Oh, sure.
Quoted 2 lines
Depending on how Xen tools will handle a PV-domain with "e820_host=1" this code might be important for domUs, too.
Right, if that's a possibility for PV (I thought that was a HVM-only thing, yet maybe it really is precisely the other way around), then yes, DomU-s may too need to cope with unexpected overlaps.
Jan
kernel test robot
build botRe: [PATCH v2 6/7] xen: allow mapping ACPI data using a different physical address
In reply to Juergen Grosskernel test robot noticed the following build errors:
Hi Juergen,
kernel test robot noticed the following build errors:
[auto build test ERROR on tip/x86/core]
[also build test ERROR on rafael-pm/linux-next rafael-pm/bleeding-edge linus/master v6.11-rc4 next-20240820] [cannot apply to xen-tip/linux-next]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]
url: https://github.com/intel-lab-lkp/linux/commits/Juergen-Gross/xen-use-correct-end-address-of-kernel-for-conflict-checking/20240820-162344 base: tip/x86/core patch link: https://lore.kernel.org/r/20240820082012.31316-7-jgross%40suse.com patch subject: [PATCH v2 6/7] xen: allow mapping ACPI data using a different physical address config: i386-buildonly-randconfig-002-20240820 (https://download.01.org/0day-ci/archive/20240821/[email protected]/config) compiler: clang version 18.1.5 (https://github.com/llvm/llvm-project 617a15a9eac96088ae5e9134248d8236e34b91b1) reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240821/[email protected]/reproduce)
If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <[email protected]>
| Closes: https://lore.kernel.org/oe-kbuild-all/[email protected]/
All errors (new ones prefixed by >>):
>> arch/x86/kernel/x86_init.c:123:23: error: use of undeclared identifier 'x86_default_set_root_pointer'
123 | .set_root_pointer = x86_default_set_root_pointer,
| ^
>> arch/x86/kernel/x86_init.c:124:23: error: use of undeclared identifier 'x86_default_get_root_pointer'
124 | .get_root_pointer = x86_default_get_root_pointer,
| ^
>> arch/x86/kernel/x86_init.c:125:28: error: use of undeclared identifier 'acpi_generic_reduced_hw_init' 125 | .reduced_hw_early_init = acpi_generic_reduced_hw_init,
| ^
3 errors generated.
vim +/x86_default_set_root_pointer +123 arch/x86/kernel/x86_init.c
f7cf5a5b8c0e59 Thomas Gleixner 2009-08-19 65
f7cf5a5b8c0e59 Thomas Gleixner 2009-08-19 66 .resources = {
5d94e81f69d4b1 Dan Williams 2011-03-08 67 .probe_roms = probe_roms,
8fee697d990c54 Thomas Gleixner 2009-08-19 68 .reserve_resources = reserve_standard_io_resources,
103e206309639b Ingo Molnar 2017-01-28 69 .memory_setup = e820__memory_setup_default,
0f4a1e80989aca Kevin Loughlin 2024-03-13 70 .dmi_setup = dmi_setup,
f7cf5a5b8c0e59 Thomas Gleixner 2009-08-19 71 },
f4848472cd9948 Thomas Gleixner 2009-08-20 72
f4848472cd9948 Thomas Gleixner 2009-08-20 73 .mpparse = {
de93410310952f Thomas Gleixner 2009-08-20 74 .setup_ioapic_ids = x86_init_noop,
e061c7ae0830ff Thomas Gleixner 2024-02-13 75 .find_mptable = mpparse_find_mptable,
dcb7600849ce9b Thomas Gleixner 2024-02-13 76 .early_parse_smp_cfg = mpparse_parse_early_smp_config,
dcb7600849ce9b Thomas Gleixner 2024-02-13 77 .parse_smp_cfg = mpparse_parse_smp_config,
f4848472cd9948 Thomas Gleixner 2009-08-20 78 },
d9112f43021554 Thomas Gleixner 2009-08-20 79
d9112f43021554 Thomas Gleixner 2009-08-20 80 .irqs = {
d9112f43021554 Thomas Gleixner 2009-08-20 81 .pre_vector_init = init_ISA_irqs,
66bcaf0bde100a Thomas Gleixner 2009-08-20 82 .intr_init = native_init_IRQ,
979923871f69a4 Thomas Gleixner 2020-01-23 83 .intr_mode_select = apic_intr_mode_select,
6b15ffa07dc325 Thomas Gleixner 2020-08-26 84 .intr_mode_init = apic_intr_mode_init,
6b15ffa07dc325 Thomas Gleixner 2020-08-26 85 .create_pci_msi_domain = native_create_pci_msi_domain,
d9112f43021554 Thomas Gleixner 2009-08-20 86 },
42bbdb43b16d23 Thomas Gleixner 2009-08-20 87
42bbdb43b16d23 Thomas Gleixner 2009-08-20 88 .oem = {
42bbdb43b16d23 Thomas Gleixner 2009-08-20 89 .arch_setup = x86_init_noop,
6f30c1ac3fcf11 Thomas Gleixner 2009-08-20 90 .banner = default_banner,
42bbdb43b16d23 Thomas Gleixner 2009-08-20 91 },
030cb6c00d242c Thomas Gleixner 2009-08-20 92
030cb6c00d242c Thomas Gleixner 2009-08-20 93 .paging = {
7737b215ad0f94 Attilio Rao 2012-08-21 94 .pagetable_init = native_pagetable_init,
030cb6c00d242c Thomas Gleixner 2009-08-20 95 },
736decac643e89 Thomas Gleixner 2009-08-19 96
736decac643e89 Thomas Gleixner 2009-08-19 97 .timers = {
736decac643e89 Thomas Gleixner 2009-08-19 98 .setup_percpu_clockev = setup_boot_APIC_clock,
845b3944bbdf9e Thomas Gleixner 2009-08-19 99 .timer_init = hpet_time_init,
c311ed6183f4fd Rahul Tanwar 2019-10-10 100 .wallclock_init = x86_wallclock_init,
736decac643e89 Thomas Gleixner 2009-08-19 101 },
d07c1be0693e09 FUJITA Tomonori 2009-11-10 102
d07c1be0693e09 FUJITA Tomonori 2009-11-10 103 .iommu = {
d07c1be0693e09 FUJITA Tomonori 2009-11-10 104 .iommu_init = iommu_init_noop,
d07c1be0693e09 FUJITA Tomonori 2009-11-10 105 },
b72d0db9dd41da Thomas Gleixner 2009-08-29 106
b72d0db9dd41da Thomas Gleixner 2009-08-29 107 .pci = {
b72d0db9dd41da Thomas Gleixner 2009-08-29 108 .init = x86_default_pci_init,
ab3b37937e8f4f Thomas Gleixner 2009-08-29 109 .init_irq = x86_default_pci_init_irq,
9325a28ce2fa7c Thomas Gleixner 2009-08-29 110 .fixup_irqs = x86_default_pci_fixup_irqs,
b72d0db9dd41da Thomas Gleixner 2009-08-29 111 },
f72e38e8ec8869 Juergen Gross 2017-11-09 112
f72e38e8ec8869 Juergen Gross 2017-11-09 113 .hyper = {
f72e38e8ec8869 Juergen Gross 2017-11-09 114 .init_platform = x86_init_noop,
f3614646005a1b Juergen Gross 2017-11-09 115 .guest_late_init = x86_init_noop,
f72e38e8ec8869 Juergen Gross 2017-11-09 116 .x2apic_available = bool_x86_init_noop,
ab0f59c6f13528 David Woodhouse 2020-10-24 117 .msi_ext_dest_id = bool_x86_init_noop,
f72e38e8ec8869 Juergen Gross 2017-11-09 118 .init_mem_mapping = x86_init_noop,
6f84f8d1587f20 Pavel Tatashin 2018-04-10 119 .init_after_bootmem = x86_init_noop,
f72e38e8ec8869 Juergen Gross 2017-11-09 120 },
038bac2b02989a Juergen Gross 2018-02-19 121
038bac2b02989a Juergen Gross 2018-02-19 122 .acpi = {
41fa1ee9c6d687 Josh Boyer 2019-08-19 @123 .set_root_pointer = x86_default_set_root_pointer,
e7b66d16fe4172 Juergen Gross 2018-10-10 @124 .get_root_pointer = x86_default_get_root_pointer,
81b53e5ff21e09 Andy Shevchenko 2018-02-20 @125 .reduced_hw_early_init = acpi_generic_reduced_hw_init,
038bac2b02989a Juergen Gross 2018-02-19 126 },
736decac643e89 Thomas Gleixner 2009-08-19 127 };736decac643e89 Thomas Gleixner 2009-08-19 128
Signature · 2 lines
0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki
kernel test robot
build botRe: [PATCH v2 6/7] xen: allow mapping ACPI data using a different physical address
In reply to Juergen Grosskernel test robot noticed the following build errors:
Hi Juergen,
kernel test robot noticed the following build errors:
[auto build test ERROR on tip/x86/core]
[also build test ERROR on rafael-pm/linux-next rafael-pm/bleeding-edge linus/master v6.11-rc4 next-20240820] [cannot apply to xen-tip/linux-next]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]
url: https://github.com/intel-lab-lkp/linux/commits/Juergen-Gross/xen-use-correct-end-address-of-kernel-for-conflict-checking/20240820-162344 base: tip/x86/core patch link: https://lore.kernel.org/r/20240820082012.31316-7-jgross%40suse.com patch subject: [PATCH v2 6/7] xen: allow mapping ACPI data using a different physical address config: x86_64-randconfig-004-20240820 (https://download.01.org/0day-ci/archive/20240821/[email protected]/config) compiler: gcc-12 (Debian 12.2.0-14) 12.2.0 reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240821/[email protected]/reproduce)
If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <[email protected]>
| Closes: https://lore.kernel.org/oe-kbuild-all/[email protected]/
All errors (new ones prefixed by >>):
In file included from arch/x86/kernel/smpboot.c:64:
>> arch/x86/include/asm/acpi.h:179:42: error: unknown type name 'acpi_physical_address'
179 | extern void __iomem * (*acpi_os_ioremap)(acpi_physical_address phys,
| ^~~~~~~~~~~~~~~~~~~~~
>> arch/x86/include/asm/acpi.h:180:42: error: unknown type name 'acpi_size'
180 | acpi_size size);
| ^~~~~~~~~
arch/x86/include/asm/acpi.h:181:35: error: unknown type name 'acpi_physical_address'
181 | void __iomem *x86_acpi_os_ioremap(acpi_physical_address phys, acpi_size size);
| ^~~~~~~~~~~~~~~~~~~~~
arch/x86/include/asm/acpi.h:181:63: error: unknown type name 'acpi_size'
181 | void __iomem *x86_acpi_os_ioremap(acpi_physical_address phys, acpi_size size);
| ^~~~~~~~~vim +/acpi_physical_address +179 arch/x86/include/asm/acpi.h
176
177 #ifdef CONFIG_XEN_PV_DOM0
178 /* A Xen PV dom0 needs a special acpi_os_ioremap() handling. */
> 179 extern void __iomem * (*acpi_os_ioremap)(acpi_physical_address phys, > 180 acpi_size size);
181 void __iomem *x86_acpi_os_ioremap(acpi_physical_address phys, acpi_size size);
182 #define acpi_os_ioremap acpi_os_ioremap
183 #endif
184
Signature · 2 lines
0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki
Juergen Gross
maintainerPATCH v2 7/7 xen: tolerate ACPI NVS memory overlapping with Xen allocated memory
In reply to Juergen GrossIn order to minimize required special handling for running as Xen PV dom0, the memory layout is modified to match that of the host. This requires to have only RAM at the locations where Xen allocated memory is living. Unfortunately there seem to be some machines, where ACPI NVS is located at 64 MB, resulting in a conflict with the loaded kernel or the initial page tables built by Xen.
As ACPI NVS needs to be accessed by the kernel only for saving and restoring it across suspend operations, it can be relocated in the dom0's memory map by swapping it with unused RAM (this is possible via modification of the dom0 P2M map).
While the E820 map can (and should) be modified right away, the P2M map can be updated only after memory allocation is working, as the P2M map might need to be extended.
---
V2:
- remap helpers split off into other patch
--- arch/x86/xen/setup.c | 92 +++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 91 insertions(+), 1 deletion(-)
arch/x86/xen/setup.c +91 -1
@@ -495,6 +495,8 @@ void __init xen_remap_memory(void)set_pte_mfn (buf, mfn_save,PAGE_KERNEL );pr_info ("Remapped %ld page(s)\n" , remapped); + +xen_do_remap_nonram (); }static unsigned long __init xen_get_pages_limit (void )
@@ -625,14 +627,102 @@ phys_addr_t __init xen_find_free_area(phys_addr_t size)return 0 ; } +/* +* Swap a non-RAM E820 map entry with RAM above ini_nr_pages. +* Note that the E820 map is modified accordingly, but the P2M map isn't yet. +* The adaption of the P2M must be deferred until page allocation is possible. +*/ +static void __init xen_e820_swap_entry_with_ram (struct e820_entry *swap_entry) +{ +struct e820_entry *entry; +unsigned int mapcnt; +phys_addr_t mem_end =PFN_PHYS (ini_nr_pages); +phys_addr_t swap_addr, swap_size, entry_end; + + swap_addr =PAGE_ALIGN_DOWN (swap_entry->addr); + swap_size =PAGE_ALIGN (swap_entry->addr - swap_addr + swap_entry->size); + entry = xen_e820_table.entries; + +for (mapcnt =0 ; mapcnt < xen_e820_table.nr_entries; mapcnt++) { + entry_end = entry->addr + entry->size; +if (entry->type ==E820_TYPE_RAM && entry->size >= swap_size && + entry_end - swap_size >= mem_end) { +/* Reduce RAM entry by needed space (whole pages). */ + entry->size -= swap_size; + +/* Add new entry at the end of E820 map. */ + entry = xen_e820_table.entries + + xen_e820_table.nr_entries; + xen_e820_table.nr_entries++; + +/* Fill new entry (keep size and page offset). */ + entry->type = swap_entry->type; + entry->addr = entry_end - swap_size + + swap_addr - swap_entry->addr; + entry->size = swap_entry->size; + +/* Convert old entry to RAM, align to pages. */ + swap_entry->type =E820_TYPE_RAM ; + swap_entry->addr = swap_addr; + swap_entry->size = swap_size; + +/* Remember PFN<->MFN relation for P2M update. */ +xen_add_remap_nonram (swap_addr, entry_end - swap_size, + swap_size); + +/* Order E820 table and merge entries. */ +e820__update_table (&xen_e820_table); + +return ; + } + + entry++; + } + +xen_raw_console_write ("No suitable area found for required E820 entry remapping action\n" ); +BUG (); +} + +/* +* Look for non-RAM memory types in a specific guest physical area and move +* those away if possible (ACPI NVS only for now). +*/ +static void __init xen_e820_resolve_conflicts (phys_addr_t start, +phys_addr_t size) +{ +struct e820_entry *entry; +unsigned int mapcnt; +phys_addr_t end; + +if (!size) +return ; + + end = start + size; + entry = xen_e820_table.entries; + +for (mapcnt =0 ; mapcnt < xen_e820_table.nr_entries; mapcnt++) { +if (entry->addr >= end) +return ; + +if (entry->addr + entry->size > start && + entry->type ==E820_TYPE_NVS ) +xen_e820_swap_entry_with_ram (entry); + + entry++; + } +} +/* * Check for an area in physical memory to be usable for non-movable purposes. -* An area is considered to usable if the used E820 map lists it to be RAM. +* An area is considered to usable if the used E820 map lists it to be RAM or +* some other type which can be moved to higher PFNs while keeping the MFNs. * In case the area is not usable, crash the system with an error message. */ void __init xen_chk_is_e820_usable (phys_addr_t start,phys_addr_t size,const char *component) { +xen_e820_resolve_conflicts (start, size); +if (!xen_is_e820_reserved (start, size))return ;
Signature · 1 line
2.43.0
Re: [PATCH v2 7/7] xen: tolerate ACPI NVS memory overlapping with Xen allocated memory
Quoted 11 lines · Juergen Gross
In order to minimize required special handling for running as Xen PV dom0, the memory layout is modified to match that of the host. This requires to have only RAM at the locations where Xen allocated memory is living. Unfortunately there seem to be some machines, where ACPI NVS is located at 64 MB, resulting in a conflict with the loaded kernel or the initial page tables built by Xen.
As ACPI NVS needs to be accessed by the kernel only for saving and restoring it across suspend operations, it can be relocated in the dom0's memory map by swapping it with unused RAM (this is possible via modification of the dom0 P2M map).
Shouldn't this paragraph be amended / re-written, as this was what made me ask for what is now patch 6?
Jan
Jürgen Groß
maintainerRe: [PATCH v2 7/7] xen: tolerate ACPI NVS memory overlapping with Xen allocated memory
Quoted 8 lines · Jan Beulich
Quoted 7 lines · Juergen Gross
In order to minimize required special handling for running as Xen PV dom0, the memory layout is modified to match that of the host. This requires to have only RAM at the locations where Xen allocated memory is living. Unfortunately there seem to be some machines, where ACPI NVS is located at 64 MB, resulting in a conflict with the loaded kernel or the initial page tables built by Xen.
Quoted 7 lines
Quoted 4 lines
As ACPI NVS needs to be accessed by the kernel only for saving and restoring it across suspend operations, it can be relocated in the dom0's memory map by swapping it with unused RAM (this is possible via modification of the dom0 P2M map).
Shouldn't this paragraph be amended / re-written, as this was what made me ask for what is now patch 6?
Oh, indeed.
Juergen