Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752767AbbF2O2Q (ORCPT ); Mon, 29 Jun 2015 10:28:16 -0400 Received: from resqmta-ch2-04v.sys.comcast.net ([69.252.207.36]:55734 "EHLO resqmta-ch2-04v.sys.comcast.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753210AbbF2O2K (ORCPT ); Mon, 29 Jun 2015 10:28:10 -0400 Date: Mon, 29 Jun 2015 09:28:08 -0500 (CDT) From: Christoph Lameter X-X-Sender: cl@east.gentwo.org To: Linus Torvalds cc: David Rientjes , linux-kernel@vger.kernel.org, Andrew Morton , Joonsoo Kim , Gavin Guo , "Theodore Ts'o" , Pekka Enberg Subject: [PATCH] Fix kmalloc slab creation sequence In-Reply-To: Message-ID: References: <20150626142332.GA28413@thunk.org> <20150627034910.GB474@thunk.org> Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4039 Lines: 104 This patch restores the slab creation sequence that was broken by commit 4066c33d0308f8 and also reverts the portions that introduced the KMALLOC_LOOP_XXX macros. Those can never really work since the slab creation is much more complex than just going from a minimum to a maximum number. The latest upstream kernel boots cleanly on my machine with a 64 bit x86 configuration under KVM using either SLAB or SLUB. Fixes: 4066c33d0308f8 ("support the slub_debug boot option") Signed-off-by: Christoph Lameter Index: linux/include/linux/slab.h =================================================================== --- linux.orig/include/linux/slab.h 2015-06-29 09:16:41.686106310 -0500 +++ linux/include/linux/slab.h 2015-06-29 09:16:41.686106310 -0500 @@ -153,30 +153,8 @@ size_t ksize(const void *); #define ARCH_KMALLOC_MINALIGN ARCH_DMA_MINALIGN #define KMALLOC_MIN_SIZE ARCH_DMA_MINALIGN #define KMALLOC_SHIFT_LOW ilog2(ARCH_DMA_MINALIGN) -/* - * The KMALLOC_LOOP_LOW is the definition for the for loop index start number - * to create the kmalloc_caches object in create_kmalloc_caches(). The first - * and the second are 96 and 192. You can see that in the kmalloc_index(), if - * the KMALLOC_MIN_SIZE <= 32, then return 1 (96). If KMALLOC_MIN_SIZE <= 64, - * then return 2 (192). If the KMALLOC_MIN_SIZE is bigger than 64, we don't - * need to initialize 96 and 192. Go directly to start the KMALLOC_SHIFT_LOW. - */ -#if KMALLOC_MIN_SIZE <= 32 -#define KMALLOC_LOOP_LOW 1 -#elif KMALLOC_MIN_SIZE <= 64 -#define KMALLOC_LOOP_LOW 2 -#else -#define KMALLOC_LOOP_LOW KMALLOC_SHIFT_LOW -#endif - #else #define ARCH_KMALLOC_MINALIGN __alignof__(unsigned long long) -/* - * The KMALLOC_MIN_SIZE of slub/slab/slob is 2^3/2^5/2^3. So, even slab is used. - * The KMALLOC_MIN_SIZE <= 32. The kmalloc-96 and kmalloc-192 should also be - * initialized. - */ -#define KMALLOC_LOOP_LOW 1 #endif /* Index: linux/mm/slab_common.c =================================================================== --- linux.orig/mm/slab_common.c 2015-06-29 09:16:41.686106310 -0500 +++ linux/mm/slab_common.c 2015-06-29 09:18:55.481973057 -0500 @@ -855,6 +855,12 @@ void __init setup_kmalloc_cache_index_ta } } +static void new_kmalloc_cache(int idx, unsigned long flags) +{ + kmalloc_caches[idx] = create_kmalloc_cache(kmalloc_info[idx].name, + kmalloc_info[idx].size, flags); +} + /* * Create the kmalloc array. Some of the regular kmalloc arrays * may already have been created because they were needed to @@ -864,25 +870,19 @@ void __init create_kmalloc_caches(unsign { int i; - for (i = KMALLOC_LOOP_LOW; i <= KMALLOC_SHIFT_HIGH; i++) { - if (!kmalloc_caches[i]) { - kmalloc_caches[i] = create_kmalloc_cache( - kmalloc_info[i].name, - kmalloc_info[i].size, - flags); - } + for (i = KMALLOC_SHIFT_LOW; i <= KMALLOC_SHIFT_HIGH; i++) { + if (!kmalloc_caches[i]) + new_kmalloc_cache(i, flags); /* - * "i == 2" is the "kmalloc-192" case which is the last special - * case for initialization and it's the point to jump to - * allocate the minimize size of the object. In slab allocator, - * the KMALLOC_SHIFT_LOW = 5. So, it needs to skip 2^3 and 2^4 - * and go straight to allocate 2^5. If the ARCH_DMA_MINALIGN is - * defined, it may be larger than 2^5 and here is also the - * trick to skip the empty gap. + * Caches that are not of the two-to-the-power-of size. + * These have to be created immediately after the + * earlier power of two caches */ - if (i == 2) - i = (KMALLOC_SHIFT_LOW - 1); + if (KMALLOC_MIN_SIZE <= 32 && !kmalloc_caches[1] && i == 6) + new_kmalloc_cache(1, flags); + if (KMALLOC_MIN_SIZE <= 64 && !kmalloc_caches[2] && i == 7) + new_kmalloc_cache(2, flags); } /* Kmalloc array is now usable */ -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/