@@ -9,6 +9,8 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/)
### Changed
- Changed flexible array definitions in public I/O interface headers to not
use "1" as the number of array elements.
+ - On x86:
+ - Reduce IOMMU setup time for hardware domain.
### Added
- On x86:
@@ -300,76 +300,6 @@ void iommu_identity_map_teardown(struct domain *d)
}
}
-static unsigned int __hwdom_init hwdom_iommu_map(const struct domain *d,
- unsigned long pfn,
- unsigned long max_pfn)
-{
- mfn_t mfn = _mfn(pfn);
- unsigned int i, type, perms = IOMMUF_readable | IOMMUF_writable;
-
- /*
- * Set up 1:1 mapping for dom0. Default to include only conventional RAM
- * areas and let RMRRs include needed reserved regions. When set, the
- * inclusive mapping additionally maps in every pfn up to 4GB except those
- * that fall in unusable ranges for PV Dom0.
- */
- if ( (pfn > max_pfn && !mfn_valid(mfn)) || xen_in_range(pfn) )
- return 0;
-
- switch ( type = page_get_ram_type(mfn) )
- {
- case RAM_TYPE_UNUSABLE:
- return 0;
-
- case RAM_TYPE_CONVENTIONAL:
- if ( iommu_hwdom_strict )
- return 0;
- break;
-
- default:
- if ( type & RAM_TYPE_RESERVED )
- {
- if ( !iommu_hwdom_inclusive && !iommu_hwdom_reserved )
- perms = 0;
- }
- else if ( is_hvm_domain(d) )
- return 0;
- else if ( !iommu_hwdom_inclusive || pfn > max_pfn )
- perms = 0;
- }
-
- /* Check that it doesn't overlap with the Interrupt Address Range. */
- if ( pfn >= 0xfee00 && pfn <= 0xfeeff )
- return 0;
- /* ... or the IO-APIC */
- if ( has_vioapic(d) )
- {
- for ( i = 0; i < d->arch.hvm.nr_vioapics; i++ )
- if ( pfn == PFN_DOWN(domain_vioapic(d, i)->base_address) )
- return 0;
- }
- else if ( is_pv_domain(d) )
- {
- /*
- * Be consistent with CPU mappings: Dom0 is permitted to establish r/o
- * ones there (also for e.g. HPET in certain cases), so it should also
- * have such established for IOMMUs.
- */
- if ( iomem_access_permitted(d, pfn, pfn) &&
- rangeset_contains_singleton(mmio_ro_ranges, pfn) )
- perms = IOMMUF_readable;
- }
- /*
- * ... or the PCIe MCFG regions.
- * TODO: runtime added MMCFG regions are not checked to make sure they
- * don't overlap with already mapped regions, thus preventing trapping.
- */
- if ( has_vpci(d) && vpci_is_mmcfg_address(d, pfn_to_paddr(pfn)) )
- return 0;
-
- return perms;
-}
-
static int __hwdom_init cf_check map_subtract(unsigned long s, unsigned long e,
void *data)
{
@@ -455,8 +385,7 @@ static int __hwdom_init cf_check identity_map(unsigned long s, unsigned long e,
void __hwdom_init arch_iommu_hwdom_init(struct domain *d)
{
- unsigned long i, top, max_pfn, start, count;
- unsigned int start_perms = 0;
+ unsigned int i;
struct rangeset *map;
struct map_data map_data = { .d = d };
int rc;
@@ -487,58 +416,46 @@ void __hwdom_init arch_iommu_hwdom_init(struct domain *d)
if ( !map )
panic("IOMMU init: unable to allocate rangeset\n");
- max_pfn = (GB(4) >> PAGE_SHIFT) - 1;
- top = max(max_pdx, pfn_to_pdx(max_pfn) + 1);
+ if ( iommu_hwdom_inclusive )
+ {
+ /* Add the whole range below 4GB, UNUSABLE regions will be removed. */
+ rc = rangeset_add_range(map, 0, PFN_DOWN(GB(4)) - 1);
+ if ( rc )
+ panic("IOMMU inclusive mappings can't be added: %d\n", rc);
+ }
- for ( i = 0, start = 0, count = 0; i < top; )
+ for ( i = 0; i < e820.nr_map; i++ )
{
- unsigned long pfn = pdx_to_pfn(i);
- unsigned int perms = hwdom_iommu_map(d, pfn, max_pfn);
+ const struct e820entry entry = e820.map[i];
- if ( !perms )
- /* nothing */;
- else if ( paging_mode_translate(d) )
+ switch ( entry.type )
{
- int rc;
-
- rc = p2m_add_identity_entry(d, pfn,
- perms & IOMMUF_writable ? p2m_access_rw
- : p2m_access_r,
- 0);
+ case E820_UNUSABLE:
+ /* Only relevant for inclusive mode, otherwise this is a no-op. */
+ rc = rangeset_remove_range(map, PFN_DOWN(entry.addr),
+ PFN_DOWN(entry.addr + entry.size - 1));
if ( rc )
- printk(XENLOG_WARNING
- "%pd: identity mapping of %lx failed: %d\n",
- d, pfn, rc);
- }
- else if ( pfn != start + count || perms != start_perms )
- {
- long rc;
+ panic("IOMMU failed to remove unusable memory: %d\n", rc);
+ continue;
- commit:
- while ( (rc = iommu_map(d, _dfn(start), _mfn(start), count,
- start_perms | IOMMUF_preempt,
- &map_data.flush_flags)) > 0 )
- {
- start += rc;
- count -= rc;
- process_pending_softirqs();
- }
- if ( rc )
- printk(XENLOG_WARNING
- "%pd: IOMMU identity mapping of [%lx,%lx) failed: %ld\n",
- d, start, start + count, rc);
- start = pfn;
- count = 1;
- start_perms = perms;
- }
- else
- ++count;
+ case E820_RESERVED:
+ if ( !iommu_hwdom_inclusive && !iommu_hwdom_reserved )
+ continue;
+ break;
- if ( !(++i & 0xfffff) )
- process_pending_softirqs();
+ case E820_RAM:
+ if ( iommu_hwdom_strict )
+ continue;
+ break;
- if ( i == top && count )
- goto commit;
+ default:
+ continue;
+ }
+
+ rc = rangeset_add_range(map, PFN_DOWN(entry.addr),
+ PFN_DOWN(entry.addr + entry.size - 1));
+ if ( rc )
+ panic("IOMMU failed to add identity range: %d\n", rc);
}
/* Remove any areas in-use by Xen. */
The current loop that iterates from 0 to the maximum RAM address in order to setup the IOMMU mappings is highly inefficient, and it will get worse as the amount of RAM increases. It's also not accounting for any reserved regions past the last RAM address. Instead of iterating over memory addresses, iterate over the memory map regions and use a rangeset in order to keep track of which ranges need to be identity mapped in the hardware domain physical address space. On an AMD EPYC 7452 with 512GiB of RAM, the time to execute arch_iommu_hwdom_init() in nanoseconds is: x old + new N Min Max Median Avg Stddev x 5 2.2364154e+10 2.338244e+10 2.2474685e+10 2.2622409e+10 4.2949869e+08 + 5 1025012 1033036 1026188 1028276.2 3623.1194 Difference at 95.0% confidence -2.26214e+10 +/- 4.42931e+08 -99.9955% +/- 9.05152e-05% (Student's t, pooled s = 3.03701e+08) Execution time of arch_iommu_hwdom_init() goes down from ~22s to ~0.001s. Note there's a change for HVM domains (ie: PVH dom0) that get switched to create the p2m mappings using map_mmio_regions() instead of p2m_add_identity_entry(), so that ranges can be mapped with a single function call if possible. Note that the interface of map_mmio_regions() doesn't allow creating read-only mappings, but so far there are no such mappings created for PVH dom0 in arch_iommu_hwdom_init(). No change intended in the resulting mappings that a hardware domain gets. Signed-off-by: Roger Pau Monné <roger.pau@citrix.com> --- Changes since v4: - Add default case to handle ACPI and NVS regions (which are not mapped) unless in the low 4GB and when inclusive mode is set. - Add changelog entry. - Dropped Jans RB. Changes since v3: - Remove unnecessary line wraps. Changes since v2: - Simplify a bit the logic related to inclusive option, at the cost of making some no-op calls on some cases. Changes since v1: - Split from bigger patch. - Remove unneeded default case. x86/iommu: add CHANGELOG entry for hwdom setup time improvements Signed-off-by: Roger Pau Monné <roger.pau@citrix.com> --- CHANGELOG.md | 2 + xen/drivers/passthrough/x86/iommu.c | 149 ++++++---------------------- 2 files changed, 35 insertions(+), 116 deletions(-)