From: John Williams <john.williams@petalogix.com>
To: monstr@seznam.cz
Cc: linux-kernel@vger.kernel.org, arnd@arndb.de,
linux-arch@vger.kernel.org, stephen.neuendorffer@xilinx.com,
John.Linn@xilinx.com, matthew@wil.cx, will.newton@gmail.com,
drepper@redhat.com, microblaze-uclinux@itee.uq.edu.au,
grant.likely@secretlab.ca, Michal Simek <monstr@monstr.eu>
Subject: Re: [PATCH 09/56] microblaze_v2: cache support
Date: Mon, 05 May 2008 12:09:03 +1000 [thread overview]
Message-ID: <1209953343.5798.54.camel@localhost> (raw)
In-Reply-To: <2f801c33caee22e112af51ae927c264ce99ead01.1209897266.git.monstr@monstr.eu>
> +++ b/arch/microblaze/kernel/cpu/cache.c
> +void __flush_icache_all(void)
> +{
> + unsigned int i;
> + unsigned flags;
> +
> + if (cpuinfo.use_icache) {
> + local_irq_save(flags);
> + __disable_icache();
> +
> + /* Just loop through cache size and invalidate, no need to add
> + CACHE_BASE address */
> + for (i = 0; i < cpuinfo.icache_size;
> + i += cpuinfo.icache_line)
> + __invalidate_icache(i);
> +
> + __enable_icache();
> + local_irq_restore(flags);
> + }
> +}
This is the slowest and greatest latency-inducing loop in the MicroBlaze
kernel. The CPU specs says icache must be disabled while you clear it -
if you have a chance it's worth checking to see what happens if you
violate that rule.
This will be less of a problem once we update the kernel_from_bram
patches, but that may be a little while yet.
> +void __flush_icache_range(unsigned long start, unsigned long end)
ditto
> +void __flush_icache_page(struct vm_area_struct *vma, struct page *page)
> +{
> + __flush_icache_all();
> +}
> +
This could be smarter - if our cache size is less than a page then there
is a win to be had. But, it's my fault so I won't complain too loudly!
> +void __flush_icache_user_range(struct vm_area_struct *vma,
> + struct page *page, unsigned long adr,
> + int len)
> +{
> + __flush_icache_all();
> +}
and again.
> +
> +void __flush_cache_sigtramp(unsigned long addr)
> +{
> + __flush_icache_range(addr, addr + 8);
> +}
Perhaps a #define in asm-microblaze/signal.h to tell us the size of the
sigtramp, and also used by microblaze/kernel/signal.c?
> +
> +void __flush_dcache_all(void)
> +{
> + unsigned int i;
> + unsigned flags;
> +
> + if (cpuinfo.use_dcache) {
> + local_irq_save(flags);
> + __disable_dcache();
> +
> + /*
> + * Just loop through cache size and invalidate,
> + * no need to add CACHE_BASE address
> + */
> + for (i = 0; i < cpuinfo.dcache_size;
> + i += cpuinfo.dcache_line)
> + __invalidate_dcache(i);
> +
> + __enable_dcache();
> + local_irq_restore(flags);
> + }
> +}
again might be worth testing if it still works with dcache enabled
during the flush loop.
> --- /dev/null
> +++ b/include/asm-microblaze/cache.h
> +#ifndef L1_CACHE_BYTES
> +/* word-granular cache in microblaze */
> +#define L1_CACHE_BYTES 4
> +#endif
Is this actually used anywhere? If so, it's wrong because cache line
sizes is configurable now.
> +/* FIXME - I don't think this is right */
> +#ifdef CONFIG_XILINX_UNCACHED_SHADOW
> +#define UNCACHED_SHADOW_MASK (CONFIG_XILINX_ERAM_SIZE)
> +#endif
If you are throwing out the alloc_consist()/free_consist()
implementation then the UNCACHED_SHADOW stuff can go too.
> +#define copy_to_user_page(vma, page, vaddr, dst, src, len) \
> +do { memcpy(dst, src, len); \
> + flush_icache_user_range(vma, page, vaddr, len); \
> +} while (0)
Is the icache flush required? Is copy_to_user_page() called from the
binfmt_flat loader?
next prev parent reply other threads:[~2008-05-05 2:09 UTC|newest]
Thread overview: 170+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-05-04 11:40 Microblaze patches V2 monstr
2008-05-04 11:40 ` [PATCH 01/56] microblaze_v2: Kconfig patches monstr
2008-05-04 11:40 ` [PATCH 02/56] microblaze_v2: Makefiles for Microblaze cpu monstr
2008-05-04 11:40 ` [PATCH 03/56] microblaze_v2: Cpuinfo handling monstr
2008-05-04 11:40 ` [PATCH 04/56] microblaze_v2: Open firmware files monstr
2008-05-04 11:40 ` [PATCH 05/56] microblaze_v2: Support for semaphores monstr
2008-05-04 11:40 ` [PATCH 06/56] microblaze_v2: exception handling monstr
2008-05-04 11:40 ` [PATCH 07/56] microblaze_v2: Signal support monstr
2008-05-04 11:40 ` [PATCH 08/56] microblaze_v2: Interrupt handling, timer support, supported function monstr
2008-05-04 11:40 ` [PATCH 09/56] microblaze_v2: cache support monstr
2008-05-04 11:40 ` [PATCH 10/56] microblaze_v2: Generic dts file for platforms monstr
2008-05-04 11:41 ` [PATCH 11/56] microblaze_v2: kernel modules support monstr
2008-05-04 11:41 ` [PATCH 12/56] microblaze_v2: lmb support monstr
2008-05-04 11:41 ` [PATCH 13/56] microblaze_v2: PVR support, cpuinfo support monstr
2008-05-04 11:41 ` [PATCH 14/56] microblaze_v2: defconfig file monstr
2008-05-04 11:41 ` [PATCH 15/56] microblaze_v2: head.S + linker script monstr
2008-05-04 11:41 ` [PATCH 16/56] microblaze_v2: supported function for memory - kernel/lib monstr
2008-05-04 11:41 ` [PATCH 17/56] microblaze_v2: checksum support monstr
2008-05-04 11:41 ` [PATCH 18/56] microblaze_v2: early_printk support monstr
2008-05-04 11:41 ` [PATCH 19/56] microblaze_v2: uaccess files monstr
2008-05-04 11:41 ` [PATCH 20/56] microblaze_v2: heartbeat file monstr
2008-05-04 11:41 ` [PATCH 21/56] microblaze_v2: setup.c - system setting monstr
2008-05-04 11:41 ` [PATCH 22/56] microblaze_v2: asm-offsets monstr
2008-05-04 11:41 ` [PATCH 23/56] microblaze_v2: process and init task function monstr
2008-05-04 11:41 ` [PATCH 24/56] microblaze_v2: time support monstr
2008-05-04 11:41 ` [PATCH 25/56] microblaze_v2: ptrace support monstr
2008-05-04 11:41 ` [PATCH 26/56] microblaze_v2: traps support monstr
2008-05-04 11:41 ` [PATCH 27/56] microblaze_v2: support for a.out monstr
2008-05-04 11:41 ` [PATCH 28/56] microblaze_v2: memory inicialization, MMU, TLB monstr
2008-05-04 11:41 ` [PATCH 29/56] microblaze_v2: page.h, segment.h, unaligned.h monstr
2008-05-04 11:41 ` [PATCH 30/56] microblaze_v2: includes SHM*, msgbuf monstr
2008-05-04 11:41 ` [PATCH 31/56] microblaze_v2: bug headers files monstr
2008-05-04 11:41 ` [PATCH 32/56] microblaze_v2: definitions of types monstr
2008-05-04 11:41 ` [PATCH 33/56] microblaze_v2: ioctl support monstr
2008-05-04 11:41 ` [PATCH 34/56] microblaze_v2: io.h IO operations monstr
2008-05-04 11:41 ` [PATCH 35/56] microblaze_v2: headers for executables format FLAT, ELF monstr
2008-05-04 11:41 ` [PATCH 36/56] microblaze_v2: dma support monstr
2008-05-04 11:41 ` [PATCH 37/56] microblaze_v2: headers for irq monstr
2008-05-04 11:41 ` [PATCH 38/56] microblaze_v2: atomic.h bitops.h byteorder.h monstr
2008-05-04 11:41 ` [PATCH 39/56] microblaze_v2: headers pgalloc.h pgtable.h monstr
2008-05-04 11:41 ` [PATCH 40/56] microblaze_v2: system.h pvr.h processor.h monstr
2008-05-04 11:41 ` [PATCH 41/56] microblaze_v2: clinkage.h linkage.h sections.h kmap_types.h monstr
2008-05-04 11:41 ` [PATCH 42/56] microblaze_v2: stats headers monstr
2008-05-04 11:41 ` [PATCH 43/56] microblaze_v2: termbits.h termios.h monstr
2008-05-04 11:41 ` [PATCH 44/56] microblaze_v2: sigcontext.h siginfo.h monstr
2008-05-04 11:41 ` [PATCH 45/56] microblaze_v2: headers simple files - empty or redirect to asm-generic monstr
2008-05-04 11:41 ` [PATCH 46/56] microblaze_v2: headers files entry.h current.h mman.h registers.h sembuf.h monstr
2008-05-04 11:41 ` [PATCH 47/56] microblaze_v2: device.h param.h topology.h monstr
2008-05-04 11:41 ` [PATCH 48/56] microblaze_v2: pool.h socket.h monstr
2008-05-04 11:41 ` [PATCH 49/56] microblaze_v2: fcntl.h sockios.h ucontext.h monstr
2008-05-04 11:41 ` [PATCH 50/56] microblaze_v2: setup.h string.h thread_info.h monstr
2008-05-04 11:41 ` [PATCH 51/56] microblaze_v2: Kbuild file monstr
2008-05-04 11:41 ` [PATCH 52/56] microblaze_v2: pci headers monstr
2008-05-04 11:41 ` [PATCH 53/56] microblaze_v2: IPC headers monstr
2008-05-04 11:41 ` [PATCH 54/56] microblaze_v2: entry.S monstr
2008-05-04 11:41 ` [PATCH 55/56] microblaze_v2: sys_microblaze.c monstr
2008-05-04 11:41 ` [PATCH 56/56] microblaze_v2: syscall_table.S and unistd.h monstr
2008-05-04 22:45 ` [PATCH 52/56] microblaze_v2: pci headers Arnd Bergmann
2008-05-05 14:08 ` Michal Simek
2008-05-05 2:39 ` [PATCH 50/56] microblaze_v2: setup.h string.h thread_info.h John Williams
2008-05-05 14:24 ` Michal Simek
2008-05-04 22:43 ` [PATCH 49/56] microblaze_v2: fcntl.h sockios.h ucontext.h Arnd Bergmann
2008-05-05 14:08 ` Michal Simek
2008-05-04 22:39 ` [PATCH 48/56] microblaze_v2: pool.h socket.h Arnd Bergmann
2008-05-04 22:37 ` [PATCH 47/56] microblaze_v2: device.h param.h topology.h Arnd Bergmann
2008-05-05 14:07 ` Michal Simek
2008-05-05 9:54 ` [PATCH 46/56] microblaze_v2: headers files entry.h current.h mman.h registers.h sembuf.h Arnd Bergmann
2008-05-06 9:07 ` Michal Simek
2008-05-06 20:57 ` Geert Uytterhoeven
2008-05-05 2:34 ` [PATCH 45/56] microblaze_v2: headers simple files - empty or redirect to asm-generic John Williams
2008-05-05 15:10 ` Grant Likely
2008-05-06 9:02 ` Arnd Bergmann
2008-05-06 14:47 ` Stephen Rothwell
2008-05-06 15:23 ` Michal Simek
2008-05-06 15:38 ` Grant Likely
2008-05-06 16:21 ` Arnd Bergmann
2008-05-07 18:46 ` Michal Simek
2008-05-05 9:50 ` [PATCH 43/56] microblaze_v2: termbits.h termios.h Arnd Bergmann
2008-05-04 22:31 ` [PATCH 42/56] microblaze_v2: stats headers Arnd Bergmann
2008-05-04 21:58 ` [PATCH 38/56] microblaze_v2: atomic.h bitops.h byteorder.h Arnd Bergmann
2008-05-05 2:28 ` John Williams
2008-05-06 8:42 ` Michal Simek
2008-05-07 7:26 ` [PATCH 37/56] microblaze_v2: headers for irq Thomas Gleixner
2008-05-11 13:56 ` Michal Simek
2008-05-05 2:25 ` [PATCH 36/56] microblaze_v2: dma support John Williams
2008-05-05 6:45 ` Geert Uytterhoeven
2008-05-06 9:16 ` Michal Simek
2008-05-06 9:48 ` Geert Uytterhoeven
2008-05-06 9:53 ` Michal Simek
2008-05-06 11:17 ` Geert Uytterhoeven
2008-05-06 11:24 ` Arnd Bergmann
2008-05-06 13:20 ` Michal Simek
2008-05-06 15:36 ` Arnd Bergmann
2008-05-07 6:24 ` Michal Simek
2008-05-07 7:17 ` Geert Uytterhoeven
2008-05-07 9:21 ` Arnd Bergmann
2008-05-07 18:43 ` Michal Simek
2008-05-05 2:24 ` [PATCH 35/56] microblaze_v2: headers for executables format FLAT, ELF John Williams
2008-05-04 21:34 ` [PATCH 33/56] microblaze_v2: ioctl support Arnd Bergmann
2008-05-05 14:06 ` Michal Simek
2008-05-04 21:28 ` [PATCH 32/56] microblaze_v2: definitions of types Arnd Bergmann
2008-05-04 21:10 ` [PATCH 30/56] microblaze_v2: includes SHM*, msgbuf Arnd Bergmann
2008-05-05 2:19 ` [PATCH 24/56] microblaze_v2: time support John Williams
2008-05-05 14:22 ` Michal Simek
2008-05-06 0:30 ` John Williams
2008-05-06 9:56 ` Michal Simek
2008-05-06 10:02 ` Michal Simek
2008-05-06 11:38 ` Arnd Bergmann
2008-05-06 13:26 ` Michal Simek
2008-05-06 22:50 ` John Williams
2008-05-06 14:28 ` Grant Likely
2008-05-06 16:36 ` Stephen Neuendorffer
2008-05-07 7:22 ` Thomas Gleixner
2008-05-05 2:15 ` [PATCH 21/56] microblaze_v2: setup.c - system setting John Williams
2008-05-05 14:21 ` Michal Simek
2008-05-05 14:36 ` [PATCH 18/56] microblaze_v2: early_printk support Grant Likely
2008-05-05 20:10 ` Michal Simek
2008-05-05 23:22 ` John Williams
2008-05-06 8:14 ` Michal Simek
2008-05-04 19:59 ` [PATCH 17/56] microblaze_v2: checksum support Arnd Bergmann
2008-05-05 14:05 ` Michal Simek
2008-05-05 2:14 ` [PATCH 13/56] microblaze_v2: PVR support, cpuinfo support John Williams
2008-05-05 14:20 ` Michal Simek
2008-05-05 2:11 ` [PATCH 12/56] microblaze_v2: lmb support John Williams
2008-05-05 21:32 ` Segher Boessenkool
2008-05-05 14:31 ` [PATCH 10/56] microblaze_v2: Generic dts file for platforms Grant Likely
2008-05-05 20:07 ` Michal Simek
2008-05-05 17:25 ` Stephen Neuendorffer
2008-05-05 23:16 ` John Williams
2008-05-05 23:32 ` Stephen Neuendorffer
2008-05-06 0:10 ` John Williams
2008-05-06 0:17 ` Stephen Neuendorffer
2008-05-06 7:50 ` Michal Simek
2008-05-06 7:38 ` Michal Simek
2008-05-05 2:09 ` John Williams [this message]
2008-05-05 17:37 ` [microblaze-uclinux] [PATCH 09/56] microblaze_v2: cache support Stephen Neuendorffer
2008-05-06 9:22 ` Michal Simek
2008-05-05 22:37 ` Stephen Neuendorffer
2008-05-05 1:58 ` [PATCH 08/56] microblaze_v2: Interrupt handling, timer support, supported function John Williams
2008-05-05 14:20 ` Michal Simek
2008-05-07 7:04 ` Thomas Gleixner
2008-05-11 13:55 ` Michal Simek
2008-05-11 14:35 ` Thomas Gleixner
2008-05-11 21:55 ` Michal Simek
2008-05-04 19:52 ` [PATCH 07/56] microblaze_v2: Signal support Arnd Bergmann
2008-05-05 21:32 ` Stephen Neuendorffer
2008-05-05 23:33 ` John Williams
2008-05-06 0:13 ` Stephen Neuendorffer
2008-05-06 0:25 ` John Williams
2008-05-06 0:33 ` Stephen Neuendorffer
2008-05-06 9:41 ` Michal Simek
2008-05-05 14:24 ` [PATCH 04/56] microblaze_v2: Open firmware files Grant Likely
2008-05-05 21:56 ` Stephen Neuendorffer
2008-05-06 7:27 ` Michal Simek
2008-05-07 16:04 ` [microblaze-uclinux] " Stephen Neuendorffer
2008-05-07 18:40 ` Michal Simek
2008-05-07 18:59 ` Grant Likely
2008-05-07 19:12 ` Michal Simek
2008-05-07 20:14 ` Stephen Neuendorffer
2008-05-05 1:52 ` [PATCH 03/56] microblaze_v2: Cpuinfo handling John Williams
2008-05-05 14:19 ` Michal Simek
2008-05-05 1:47 ` [PATCH 02/56] microblaze_v2: Makefiles for Microblaze cpu John Williams
2008-05-06 8:51 ` Michal Simek
2008-05-04 21:24 ` [PATCH 01/56] microblaze_v2: Kconfig patches Grant Likely
2008-05-05 6:36 ` Michal Simek
2008-05-05 1:42 ` John Williams
2008-05-05 6:46 ` Michal Simek
2008-05-05 14:16 ` Michal Simek
2008-05-05 2:30 ` Microblaze patches V2 John Williams
2008-05-05 7:02 ` Michal Simek
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1209953343.5798.54.camel@localhost \
--to=john.williams@petalogix.com \
--cc=John.Linn@xilinx.com \
--cc=arnd@arndb.de \
--cc=drepper@redhat.com \
--cc=grant.likely@secretlab.ca \
--cc=linux-arch@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=matthew@wil.cx \
--cc=microblaze-uclinux@itee.uq.edu.au \
--cc=monstr@monstr.eu \
--cc=monstr@seznam.cz \
--cc=stephen.neuendorffer@xilinx.com \
--cc=will.newton@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).