[U-Boot] [PATCH 01/23] x86: Enhance the microcode tool to support header files as input
Simon Glass
sjg at chromium.org
Tue Jan 27 16:12:32 CET 2015
Hi Bin,
On 27 January 2015 at 01:59, Bin Meng <bmeng.cn at gmail.com> wrote:
> Hi Simon,
>
> On Tue, Jan 27, 2015 at 9:22 AM, Simon Glass <sjg at chromium.org> wrote:
>> Sometimes microcode is delivered as a header file. Allow the tool to
>> support this as well as collecting multiple microcode blocks into a
>> single update.
>>
>> Signed-off-by: Simon Glass <sjg at chromium.org>
>> ---
>
> Tested-by: Bin Meng <bmeng.cn at gmail.com>
>
> But please see my comments below.
>
>> tools/microcode-tool.py | 90 ++++++++++++++++++++++++++++++++++++++-----------
>> 1 file changed, 70 insertions(+), 20 deletions(-)
>>
>> diff --git a/tools/microcode-tool.py b/tools/microcode-tool.py
>> index 003716d..71c2e91 100755
>> --- a/tools/microcode-tool.py
>> +++ b/tools/microcode-tool.py
>> @@ -76,6 +76,35 @@ def ParseFile(fname):
>> microcodes[name] = Microcode(name, data)
>> return date, license_text, microcodes
>>
>> +def ParseHeaderFiles(fname_list):
>> + """Parse a list of header files and return the component parts
>> +
>> + Args:
>> + fname_list: List of files to parse
>> + Returns:
>> + date: String containing date from the file's header
>> + license_text: List of text lines for the license file
>> + microcodes: List of Microcode objects from the file
>> + """
>> + microcodes = {}
>> + license_text = []
>> + date = ''
>> + name = None
>> + for fname in fname_list:
>> + name = os.path.basename(fname).lower()
>> + name = os.path.splitext(name)[0]
>> + data = []
>> + with open(fname) as fd:
>> + for line in fd:
>> + line = line.rstrip()
>> +
>> + # Omit anything after the last comma
>> + words = line.split(',')[:-1]
>> + data += [word + ',' for word in words]
>> + microcodes[name] = Microcode(name, data)
>> + return date, license_text, microcodes
>> +
>> +
>> def List(date, microcodes, model):
>> """List the available microcode chunks
>>
>> @@ -129,13 +158,13 @@ def FindMicrocode(microcodes, model):
>> break
>> return found, tried
>>
>> -def CreateFile(date, license_text, mcode, outfile):
>> +def CreateFile(date, license_text, mcodes, outfile):
>> """Create a microcode file in U-Boot's .dtsi format
>>
>> Args:
>> date: String containing date of original microcode file
>> license: List of text lines for the license file
>> - mcode: Microcode object to write
>> + mcodes: Microcode objects to write (normally only 1)
>> outfile: Filename to write to ('-' for stdout)
>> """
>> out = '''/*%s
>> @@ -159,15 +188,22 @@ intel,processor-flags = <%#x>;
>> data = <%s
>> \t>;'''
>> words = ''
>> - for i in range(len(mcode.words)):
>> - if not (i & 3):
>> - words += '\n'
>> - val = mcode.words[i]
>> - # Change each word so it will be little-endian in the FDT
>> - # This data is needed before RAM is available on some platforms so we
>> - # cannot do an endianness swap on boot.
>> - val = struct.unpack("<I", struct.pack(">I", val))[0]
>> - words += '\t%#010x' % val
>> + add_comments = len(mcodes) > 1
>> + for mcode in mcodes:
>> + if add_comments:
>> + words += '\n/* %s */' % mcode.name
>> + for i in range(len(mcode.words)):
>> + if not (i & 3):
>> + words += '\n'
>> + val = mcode.words[i]
>> + # Change each word so it will be little-endian in the FDT
>> + # This data is needed before RAM is available on some platforms so
>> + # we cannot do an endianness swap on boot.
>> + val = struct.unpack("<I", struct.pack(">I", val))[0]
>> + words += '\t%#010x' % val
>> +
>> + # Use the first microcode for the headers
>> + mcode = mcodes[0]
>>
>> # Take care to avoid adding a space before a tab
>> text = ''
>> @@ -187,8 +223,8 @@ data = <%s
>> print >> sys.stderr, "Creating directory '%s'" % MICROCODE_DIR
>> os.makedirs(MICROCODE_DIR)
>> outfile = os.path.join(MICROCODE_DIR, mcode.name + '.dtsi')
>> - print >> sys.stderr, "Writing microcode for '%s' to '%s'" % (
>> - mcode.name, outfile)
>> + print >> sys.stderr, "Writing microcode for '%s' to '%s'" % (
>> + ', '.join([mcode.name for mcode in mcodes]), outfile)
>> with open(outfile, 'w') as fd:
>> print >> fd, out % tuple(args)
>>
>> @@ -198,8 +234,12 @@ def MicrocodeTool():
>> parser = OptionParser()
>> parser.add_option('-d', '--mcfile', type='string', action='store',
>> help='Name of microcode.dat file')
>> + parser.add_option('-H', '--headerfile', type='string', action='append',
>> + help='Name of .h file containing microcode')
>> parser.add_option('-m', '--model', type='string', action='store',
>> - help='Model name to extract')
>> + help="Model name to extract ('all' for all)")
>> + parser.add_option('-M', '--multiple', type='string', action='store',
>> + help="Allow output of multiple models")
>> parser.add_option('-o', '--outfile', type='string', action='store',
>> help='Filename to use for output (- for stdout), default is'
>> ' %s/<name>.dtsi' % MICROCODE_DIR)
>> @@ -224,9 +264,14 @@ def MicrocodeTool():
>> if cmd not in commands:
>> parser.error("Unknown command '%s'" % cmd)
>>
>> - if not options.mcfile:
>> - parser.error('You must specify a microcode file')
>> - date, license_text, microcodes = ParseFile(options.mcfile)
>> + if (not not options.mcfile) != (not not options.mcfile):
>> + parser.error("You must specify either header files or a microcode file, not both")
>> + if options.headerfile:
>> + date, license_text, microcodes = ParseHeaderFiles(options.headerfile)
>> + elif options.mcfile:
>> + date, license_text, microcodes = ParseFile(options.mcfile)
>> + else:
>> + parser.error('You must specify a microcode file (or header files)')
>>
>> if cmd == 'list':
>> List(date, microcodes, options.model)
>> @@ -236,16 +281,21 @@ def MicrocodeTool():
>> if not options.model:
>> parser.error('You must specify a model to create')
>> model = options.model.lower()
>> - mcode_list, tried = FindMicrocode(microcodes, model)
>> + if options.model == 'all':
>> + options.multiple = True
>> + mcode_list = microcodes.values()
>> + tried = []
>> + else:
>> + mcode_list, tried = FindMicrocode(microcodes, model)
>> if not mcode_list:
>> parser.error("Unknown model '%s' (%s) - try 'list' to list" %
>> (model, ', '.join(tried)))
>> - if len(mcode_list) > 1:
>> + if not options.multiple and len(mcode_list) > 1:
>> parser.error("Ambiguous model '%s' (%s) matched %s - try 'list' "
>> "to list or specify a particular file" %
>> (model, ', '.join(tried),
>> ', '.join([m.name for m in mcode_list])))
>> - CreateFile(date, license_text, mcode_list[0], options.outfile)
>> + CreateFile(date, license_text, mcode_list, options.outfile)
>> else:
>> parser.error("Unknown command '%s'" % cmd)
>>
>> --
>
> My testing results show that, compared to '-d microcode.dat' approach,
> generated microcode.dtsi using '-H' does not have license header and
> date set. Is this intentional?
Not really, but the header I tested with has no license info.
I suspect we could enable use of -d also to pick this up. But we still
wouldn't get the date. I'm not sure what automation we should have for
this case.
Regards,
Simon
More information about the U-Boot
mailing list