linux-kernel - Re: [PATCH 01/56] microblaze

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Mon, 05 May 2008 16:16:45 +0200
From:	Michal Simek <monstr@...nam.cz>
To:	John Williams <john.williams@...alogix.com>
CC:	linux-kernel@...r.kernel.org, arnd@...db.de,
	linux-arch@...r.kernel.org, stephen.neuendorffer@...inx.com,
	John.Linn@...inx.com, matthew@....cx, will.newton@...il.com,
	drepper@...hat.com, microblaze-uclinux@...e.uq.edu.au,
	grant.likely@...retlab.ca, Michal Simek <monstr@...str.eu>
Subject: Re: [PATCH 01/56] microblaze_v2: Kconfig patches

Hi John

>> +config EARLY_PRINTK_UARTLITE_ADDRESS
>> +	hex "Physical address where UART Lite for early printk is mapped"
>> +	depends on EARLY_PRINTK
>> +	default "0x40100000"
>> +	help
>> +	  Please enter physical address where your uart lite is mapped.
> 
> I think this should default to 0x00000000, then in the early_printk code
> we can test for zero baseaddress and attempt some sensible fallback in
> that case (even to disable).  

FIXED and use EARLY_PRINTK to default set to no.

>> +
>> +config HEART_BEAT
>> +	bool "Heart beat function for kernel"
>> +	help
>> +	  This option turns on/off heart beat kernel functionality on selected
>> +	  GPIO address.
>> +
>> +config HEART_BEAT_ADDRESS
>> +	hex "Physical address for heart beat LED"
>> +	depends on HEART_BEAT
>> +	default "0x40600000"
>> +	help
>> +	  Please enter physical address GPIO LED.
> 
> Again, default address should be zero which is checked in the code (zero
> can disable).  Prevent accesses to random memory locations in case of
> poorly configured kernel.

the same

>> +config HACK
>> +	bool "Increase performance"
>> +	default y
> 
> As suggested by Grant, this sohuld be
> 
> 1. renamed and described as something sensible
> 
> SELF_MODIFYING_CODE_HACK
> 
> 2. Made condition on CONFIG_EXPERIMENTAL at very least.


  21 config SELFMOD
  22         bool "Use self modified code for intc/timer"
  23         depends on EXPERIMENTAL
  24         default n
  25         help
  26           This choice enables self-modified code for interrupt controller
  27           and timer.
  28
  29 config SELFMOD_INTC
  30         bool "Use self modified code for intc"
  31         depends on SELFMOD
  32         default y
  33         help
  34           This choice enables self-modified code for interrupt controller.
  35
  36 config SELFMOD_TIMER
  37         bool "Use self modified code for timer"
  38         depends on SELFMOD
  39         default y
  40         help
  41           This choice enables self-modified code for timer.

FIXED.

>> +config OPT_LIB_FUNCTION
> +	bool "Optimalized lib function"
>> +	default y
>> +	help
>> +	  Allows turn on optimalized library function (memcpy and memmove).
>> +	  They are optimized by using unsigned alignment. This will work
> 
> "Unsigned alignment"??
> 
> Do you mean word-aligned?

FIXED

>> +	  fine if both source and destination are aligned on the same
>> +	  boundary. However, if they are aligned on different boundaries
>> +	  shifts will be necessary. This might result in bad performance
>> +	  on MicroBlaze systems without a barrel shifter.
>> +
>> +# This is still a bit broken - disabling for now JW 20070504
>> +config ALLOW_EDIT_AUTO
>> +	bool "Permit Display/edit of Kconfig.auto platform settings"
>> +	default n
>> +	help
>> +	  Allows the editing of auto-generated platform settings from
>> +	  the Kconfig.auto file. Obviously this does not change the
>> +	  underlying hardware, so be very careful if you go editing
>> +	  these settings.
>> +
>> +	  Also, if you enable this, and edit various Kconfig.auto
>> +	  settings, YOUR CHANGES WILL BE LOST if you then disable it
>> +	  again. You have been warned!
>> +
>> +	  If unsure, say no.
>> +
>> +# Ok, the platform is chosen. Source the kconfig.auto to get all of the
>> +# system settings.
> 
> Should be just "get all of the CPU settings", since we use OF to get the
> device tree.

I keep this for now. I hope we will talked about later with Xilinx guys what
happen with platforms.

>> diff --git a/arch/microblaze/platform/generic/Kconfig.auto b/arch/microblaze/platform/generic/Kconfig.auto
>> new file mode 100644
>> index 0000000..5dc8c15
>> --- /dev/null
>> +++ b/arch/microblaze/platform/generic/Kconfig.auto
> 
>> +config XILINX_ERAM_SIZE
>> +	hex "Memory size of XILINX_RAM"
>> +	depends on XILINX_UNCACHED_SHADOW
>> +	default 0x02000000
> 
> Why does this depend on UNCACHED_SHADOW?

Because consistent.c needs to know where was the end of memory. This issue
hasn't never tested. Currently consistent.c is removed.

M
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/