Author: markj
Date: Mon Aug 31 14:47:23 2020
New Revision: 364988
URL: https://svnweb.freebsd.org/changeset/base/364988
Log:
Stop accounting for MAX_KMAP when allocating UMA boot pages.
With r364987 kernel maps are allocated statically. In head, a different
mechanism is used to bootstrap UMA.
This is a direct commit to stable/12.
Reported by: Jenkins
Modified:
stable/12/sys/vm/vm_page.c
Modified: stable/12/sys/vm/vm_page.c
==============================================================================
--- stable/12/sys/vm/vm_page.c Mon Aug 31 14:23:41 2020 (r364987)
+++ stable/12/sys/vm/vm_page.c Mon Aug 31 14:47:23 2020 (r364988)
@@ -590,9 +590,6 @@ vm_page_startup(vm_offset_t vaddr)
#ifndef UMA_MD_SMALL_ALLOC
/* vmem_startup() calls uma_prealloc(). */
boot_pages += vmem_startup_count();
- /* vm_map_startup() calls uma_prealloc(). */
- boot_pages += howmany(MAX_KMAP,
- UMA_SLAB_SPACE / sizeof(struct vm_map));
/*
* Before going fully functional kmem_init() does allocation
_______________________________________________
[email protected] mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "[email protected]"
