Date: Mon, 29 Jul 2019 21:21:54 +0000 (UTC) From: Mark Johnston <markj@FreeBSD.org> To: src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-head@freebsd.org Subject: svn commit: r350427 - head/sys/arm64/arm64 Message-ID: <201907292121.x6TLLsiX022787@repo.freebsd.org>
next in thread | raw e-mail | index | archive | help
Author: markj Date: Mon Jul 29 21:21:53 2019 New Revision: 350427 URL: https://svnweb.freebsd.org/changeset/base/350427 Log: Have arm64's pmap_fault() handle WnR faults on dirty PTEs. If we take a WnR permission fault on a managed, writeable and dirty PTE, simply return success without calling the main fault handler. This situation can occur if multiple threads simultaneously access a clean writeable mapping and trigger WnR faults; losers of the race to mark the PTE dirty would end up calling the main fault handler, which had no work to do. Reported by: alc Reviewed by: alc MFC with: r350004 Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D21097 Modified: head/sys/arm64/arm64/pmap.c Modified: head/sys/arm64/arm64/pmap.c ============================================================================== --- head/sys/arm64/arm64/pmap.c Mon Jul 29 20:50:26 2019 (r350426) +++ head/sys/arm64/arm64/pmap.c Mon Jul 29 21:21:53 2019 (r350427) @@ -5743,7 +5743,7 @@ pmap_sync_icache(pmap_t pmap, vm_offset_t va, vm_size_ int pmap_fault(pmap_t pmap, uint64_t esr, uint64_t far) { - pt_entry_t *pte; + pt_entry_t pte, *ptep; register_t intr; uint64_t ec, par; int lvl, rv; @@ -5767,9 +5767,9 @@ pmap_fault(pmap_t pmap, uint64_t esr, uint64_t far) case ISS_DATA_DFSC_AFF_L2: case ISS_DATA_DFSC_AFF_L3: PMAP_LOCK(pmap); - pte = pmap_pte(pmap, far, &lvl); - if (pte != NULL) { - pmap_set_bits(pte, ATTR_AF); + ptep = pmap_pte(pmap, far, &lvl); + if (ptep != NULL) { + pmap_set_bits(ptep, ATTR_AF); rv = KERN_SUCCESS; /* * XXXMJ as an optimization we could mark the entry @@ -5785,12 +5785,13 @@ pmap_fault(pmap_t pmap, uint64_t esr, uint64_t far) (esr & ISS_DATA_WnR) == 0) return (rv); PMAP_LOCK(pmap); - pte = pmap_pte(pmap, far, &lvl); - if (pte != NULL && - (pmap_load(pte) & (ATTR_AP_RW_BIT | ATTR_SW_DBM)) == - (ATTR_AP(ATTR_AP_RO) | ATTR_SW_DBM)) { - pmap_clear_bits(pte, ATTR_AP_RW_BIT); - pmap_invalidate_page(pmap, far); + ptep = pmap_pte(pmap, far, &lvl); + if (ptep != NULL && + ((pte = pmap_load(ptep)) & ATTR_SW_DBM) != 0) { + if ((pte & ATTR_AP_RW_BIT) == ATTR_AP(ATTR_AP_RO)) { + pmap_clear_bits(ptep, ATTR_AP_RW_BIT); + pmap_invalidate_page(pmap, far); + } rv = KERN_SUCCESS; } PMAP_UNLOCK(pmap);
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201907292121.x6TLLsiX022787>