From mboxrd@z Thu Jan 1 00:00:00 1970 From: Peter Zijlstra Date: Thu, 03 Sep 2009 08:46:30 +0000 Subject: RE: [RFC PATCH] Add TRACE_IRQFLAGS_SUPPORT, LOCKDEP_SUPPORT then Message-Id: <1251967590.7447.0.camel@twins> List-Id: References: <3877989d0908272018va6ee8f7n7ab2838fdf1b26e6@mail.gmail.com> <20090828062212.GB11552@elte.hu> <57C9024A16AD2D4C97DC78E552063EA3E038DE7E@orsmsx505.amr.corp.intel.com> In-Reply-To: <57C9024A16AD2D4C97DC78E552063EA3E038DE7E@orsmsx505.amr.corp.intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: "Luck, Tony" Cc: Ingo Molnar , Luming Yu , Christoph Hellwig , LKML , "linux-ia64@vger.kernel.org" , "Yu, Fenghua" , Felix Blyakher , "Li, Shaohua" , Bob Picco On Wed, 2009-09-02 at 16:27 -0700, Luck, Tony wrote: > >> diff -BruN linux-2.6.31-rc6/arch/ia64/include/asm/page.h > >> linux-2.6.31-rc6-lockdep/arch/ia64/include/asm/page.h > >> --- linux-2.6.31-rc6/arch/ia64/include/asm/page.h 2009-08-13 > >> 15:43:34.000000000 -0700 > >> +++ linux-2.6.31-rc6-lockdep/arch/ia64/include/asm/page.h 2009-08-23 > >> 18:59:00.000000000 -0700 > >> @@ -41,7 +41,7 @@ > >> #define PAGE_SIZE (__IA64_UL_CONST(1) << PAGE_SHIFT) > >> #define PAGE_MASK (~(PAGE_SIZE - 1)) > >> > >> -#define PERCPU_PAGE_SHIFT 16 /* log2() of max. size of per-CPU area */ > >> +#define PERCPU_PAGE_SHIFT 20 /*16 log2() of max. size of per-CPU area */ > >> #define PERCPU_PAGE_SIZE (__IA64_UL_CONST(1) << PERCPU_PAGE_SHIFT) > > > > Why was this seemingly unrelated change done in a lockdep patch? > > IIRC it is because the lockdep code needs a lot more percpu memory than the > 64K that ia64 allocates. > > But I think that I'd prefer to only allocate the extra space when it > is needed: > > #ifdef CONFIG_LOCKDEP > #define PERCPU_PAGE_SHIFT 20 /* Need a Megabyte for lockdep */ > #else > #define PERCPU_PAGE_SHIFT 16 /* 64K should be enough for normal use */ > #endif Hmm, isn't all that percpu stuff in a special section, who's size you can query at boot time and set-up accordingly?