Date: Sun, 17 Dec 2017 04:19:19 +0000 (UTC) From: Jeff Roberson <jeff@FreeBSD.org> To: src-committers@freebsd.org, svn-src-user@freebsd.org Subject: svn commit: r326917 - user/jeff/numa/sys/kern Message-ID: <201712170419.vBH4JJ1l073982@repo.freebsd.org>
next in thread | raw e-mail | index | archive | help
Author: jeff Date: Sun Dec 17 04:19:19 2017 New Revision: 326917 URL: https://svnweb.freebsd.org/changeset/base/326917 Log: Use the correct KVA domain when allocating vmem boundary tags so that we don't end up mixing super-page and non-super-page allocations. Modified: user/jeff/numa/sys/kern/subr_vmem.c Modified: user/jeff/numa/sys/kern/subr_vmem.c ============================================================================== --- user/jeff/numa/sys/kern/subr_vmem.c Sun Dec 17 04:17:36 2017 (r326916) +++ user/jeff/numa/sys/kern/subr_vmem.c Sun Dec 17 04:19:19 2017 (r326917) @@ -625,15 +625,15 @@ vmem_bt_alloc(uma_zone_t zone, vm_size_t bytes, int do * and memory are added in one atomic operation. */ mtx_lock(&vmem_bt_lock); - if (vmem_xalloc(kernel_arena, bytes, 0, 0, 0, VMEM_ADDR_MIN, - VMEM_ADDR_MAX, M_NOWAIT | M_NOVM | M_USE_RESERVE | M_BESTFIT, - &addr) == 0) { + if (vmem_xalloc(vm_dom[domain].vmd_kernel_arena, bytes, 0, 0, 0, + VMEM_ADDR_MIN, VMEM_ADDR_MAX, + M_NOWAIT | M_NOVM | M_USE_RESERVE | M_BESTFIT, &addr) == 0) { if (kmem_back_domain(domain, kernel_object, addr, bytes, M_NOWAIT | M_USE_RESERVE) == 0) { mtx_unlock(&vmem_bt_lock); return ((void *)addr); } - vmem_xfree(kernel_arena, addr, bytes); + vmem_xfree(vm_dom[domain].vmd_kernel_arena, addr, bytes); mtx_unlock(&vmem_bt_lock); /* * Out of memory, not address space. This may not even be
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201712170419.vBH4JJ1l073982>