From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Tue, 8 Apr 2003 14:56:45 -0700 From: Matt Porter To: invictus rm Cc: porter@cox.net, linuxppc-dev@lists.linuxppc.org Subject: Re: insmod search_exception_table !!! Message-ID: <20030408145645.A21872@home.com> References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: ; from invictus_rm@hotmail.com on Tue, Apr 08, 2003 at 10:08:35PM +0530 Sender: owner-linuxppc-dev@lists.linuxppc.org List-Id: On Tue, Apr 08, 2003 at 10:08:35PM +0530, invictus rm wrote: > Hi Matt, > I am using io_block_mapping for early_console_prints > io_block_mapping(0xf0000000, 0xf0000000, 0x10000000, _PAGE_IO); > > Is this a cause of the problem ?? Not on its own...you should have ~240MB of initial vmalloc space with this configuration. > How do i know how much addr i have ioremapped ?? You really want to know how much vmalloc space has been consumed by all users. Hard way: Count all the ioremaps you do (plus 4KB offset between each vmalloc offset). If you are on a non-coherent platform you also have to consider all consistent alloced memory. Easy way: ftp://source.mvista.com/pub/linuxppc/vmalloc_stats-2_4_devel.patch /proc/meminfo contains vmalloc usage stats. > My PAGE_OFFSET is at 0xc0000000 > > Do i need to change the PCI MEM Map also if i am having a 1:1 mapping in the > kernel virtual space ?? No, you just need to be aware that VMALLOC_END=ioremap_bot and this value is set to whatever your lowest io_block_map virtual address is. Any early ioremaps (pre-MM initialized) you do to place system devices will further drop VMALLOC_END and constrain your vmalloc space. Regards, -- Matt Porter porter@cox.net This is Linux Country. On a quiet night, you can hear Windows reboot. ** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/