From patchwork Thu May 16 14:01:25 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Wilco Dijkstra X-Patchwork-Id: 32707 Received: (qmail 108376 invoked by alias); 16 May 2019 14:01:31 -0000 Mailing-List: contact libc-alpha-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Unsubscribe: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: libc-alpha-owner@sourceware.org Delivered-To: mailing list libc-alpha@sourceware.org Received: (qmail 108360 invoked by uid 89); 16 May 2019 14:01:31 -0000 Authentication-Results: sourceware.org; auth=none X-Spam-SWARE-Status: No, score=-17.1 required=5.0 tests=AWL, BAYES_00, GIT_PATCH_0, GIT_PATCH_1, GIT_PATCH_2, GIT_PATCH_3, RCVD_IN_DNSWL_NONE, SPF_HELO_PASS, SPF_PASS autolearn=ham version=3.3.1 spammy= X-HELO: EUR03-DB5-obe.outbound.protection.outlook.com DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=armh.onmicrosoft.com; s=selector2-armh-onmicrosoft-com; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-SenderADCheck; bh=g1zGfbuwa0pLDWj/PMOVYEC30Pd4TNw74KYBAOPQ5SE=; b=WuT36x7uB5ZJnNGQ/DwELUmKE24uK9r5xooQLyHEzfVzryJ7ja+lvP9WPvSXy4W/TziAVOsbVbd5zZk0sakp5m5zcdIT67gCICMidSWCS2jndwLlWNmMmWgM3P7g24NmTs1OeFa4YMIAH7u2Za85RGIRExmbuupG+qAVpb/sQ4k= From: Wilco Dijkstra To: DJ Delorie CC: "libc-alpha@sourceware.org" , nd Subject: Re: [PATCH] Small tcache improvements Date: Thu, 16 May 2019 14:01:25 +0000 Message-ID: References: (message from Wilco Dijkstra on Mon, 13 May 2019 13:13:02 +0000), In-Reply-To: authentication-results: spf=none (sender IP is ) smtp.mailfrom=Wilco.Dijkstra@arm.com; x-ms-oob-tlc-oobclassifiers: OLM:3276; received-spf: None (protection.outlook.com: arm.com does not designate permitted sender hosts) x-ms-exchange-senderadcheck: 1 MIME-Version: 1.0 X-MS-Exchange-CrossTenant-mailboxtype: HOSTED Hi DJ, > It's "the GNU way" to avoid arbitrary limitations when possible, at > least it's a subtle hint mentioned elsewhere: Sure, here is the updated patch with limit set to UINT16_MAX: Change the tcache->counts[] entries to uint16_t - this removes the limit set by char and allows a larger tcache. bench-malloc-thread with 4 threads is ~15% faster on Cortex-A72. OK for commit? ChangeLog: 2019-05-16 Wilco Dijkstra * malloc/malloc.c (MAX_TCACHE_COUNT): Increase to UINT16_MAX. (tcache_put): Remove redundant assert. (tcache_get): Remove redundant asserts. (__libc_malloc): Check tcache count is not zero. * manual/tunables.texi (glibc.malloc.tcache_count): Update maximum. diff --git a/malloc/malloc.c b/malloc/malloc.c index b8baaa2706d8d274b04b86e27fc72716753530b0..00ce48cf5879c87f051af781e9b2d99f384e55dd 100644 --- a/malloc/malloc.c +++ b/malloc/malloc.c @@ -321,6 +321,10 @@ __malloc_assert (const char *assertion, const char *file, unsigned int line, /* This is another arbitrary limit, which tunables can change. Each tcache bin will hold at most this number of chunks. */ # define TCACHE_FILL_COUNT 7 + +/* Maximum chunks in tcache bins for tunables. This value must fit the range + of tcache->counts[] entries, else they may overflow. */ +# define MAX_TCACHE_COUNT UINT16_MAX #endif @@ -2901,12 +2905,10 @@ typedef struct tcache_entry time), this is for performance reasons. */ typedef struct tcache_perthread_struct { - char counts[TCACHE_MAX_BINS]; + uint16_t counts[TCACHE_MAX_BINS]; tcache_entry *entries[TCACHE_MAX_BINS]; } tcache_perthread_struct; -#define MAX_TCACHE_COUNT 127 /* Maximum value of counts[] entries. */ - static __thread bool tcache_shutting_down = false; static __thread tcache_perthread_struct *tcache = NULL; @@ -2916,7 +2918,6 @@ static __always_inline void tcache_put (mchunkptr chunk, size_t tc_idx) { tcache_entry *e = (tcache_entry *) chunk2mem (chunk); - assert (tc_idx < TCACHE_MAX_BINS); /* Mark this chunk as "in the tcache" so the test in _int_free will detect a double free. */ @@ -2933,8 +2934,6 @@ static __always_inline void * tcache_get (size_t tc_idx) { tcache_entry *e = tcache->entries[tc_idx]; - assert (tc_idx < TCACHE_MAX_BINS); - assert (tcache->counts[tc_idx] > 0); tcache->entries[tc_idx] = e->next; --(tcache->counts[tc_idx]); e->key = NULL; @@ -3046,9 +3045,8 @@ __libc_malloc (size_t bytes) DIAG_PUSH_NEEDS_COMMENT; if (tc_idx < mp_.tcache_bins - /*&& tc_idx < TCACHE_MAX_BINS*/ /* to appease gcc */ && tcache - && tcache->entries[tc_idx] != NULL) + && tcache->counts[tc_idx] > 0) { return tcache_get (tc_idx); } diff --git a/manual/tunables.texi b/manual/tunables.texi index ae638823a21b9cc7aca3684c8e3067cb8cd287e0..ee0fdf28d11b1f375b89ab0d2e011ef78170b5c9 100644 --- a/manual/tunables.texi +++ b/manual/tunables.texi @@ -189,7 +189,7 @@ per-thread cache. The default (and maximum) value is 1032 bytes on @deftp Tunable glibc.malloc.tcache_count The maximum number of chunks of each size to cache. The default is 7. -The upper limit is 127. If set to zero, the per-thread cache is effectively +The upper limit is 65535. If set to zero, the per-thread cache is effectively disabled. The approximate maximum overhead of the per-thread cache is thus equal