Comments for The Shape of Code http://shape-of-code.coding-guidelines.com Fri, 12 Mar 2010 00:00:07 +0000 http://wordpress.org/?v=2.6 Comment on Designing a processor for increased source portability costs by Derek-Jones http://shape-of-code.coding-guidelines.com/2010/02/designing-a-processor-for-increased-source-portability-costs/#comment-1517 Derek-Jones Tue, 09 Feb 2010 12:43:00 +0000 http://shape-of-code.coding-guidelines.com/?p=177#comment-1517 <a href="#comment-1490" rel="nofollow">@cate </a> Thanks for pointing out that I need to discuss 8-bit chars; now done! The Intel x86 and many RISC chips have opposite byte endianness which ensures that this issue is addressed by a lot of Open Source code. What about word endianness (i.e., swapping 16-bit quantities) like that used by the DEC (now HP) <a href="http://en.wikipedia.org/wiki/Vax" rel="nofollow">VAX</a>? This is an interesting possibility that I had overlooked. Having pointers occupy more bits than a long (the IBM RS/400 has 16 byte pointers!) introduces all sorts of extra costs for the processor vendor, so I don't think that this is a cost effective option. @cate Thanks for pointing out that I need to discuss 8-bit chars; now done!

The Intel x86 and many RISC chips have opposite byte endianness which ensures that this issue is addressed by a lot of Open Source code. What about word endianness (i.e., swapping 16-bit quantities) like that used by the DEC (now HP) VAX? This is an interesting possibility that I had overlooked.

Having pointers occupy more bits than a long (the IBM RS/400 has 16 byte pointers!) introduces all sorts of extra costs for the processor vendor, so I don’t think that this is a cost effective option.

]]> Comment on Designing a processor for increased source portability costs by Derek-Jones http://shape-of-code.coding-guidelines.com/2010/02/designing-a-processor-for-increased-source-portability-costs/#comment-1513 Derek-Jones Tue, 09 Feb 2010 12:01:07 +0000 http://shape-of-code.coding-guidelines.com/?p=177#comment-1513 <a href="#comment-1482" rel="nofollow">@Tony Finch </a> The diagram on that page clearly shows zero at the bottom and FFF... at the top. In a signed address space zero sits in the middle of memory. I did not see any obvious errors in the article saying that the address space was signed. @Tony Finch
The diagram on that page clearly shows zero at the bottom and FFF… at the top. In a signed address space zero sits in the middle of memory. I did not see any obvious errors in the article saying that the address space was signed.

]]> Comment on Designing a processor for increased source portability costs by cate http://shape-of-code.coding-guidelines.com/2010/02/designing-a-processor-for-increased-source-portability-costs/#comment-1490 cate Tue, 09 Feb 2010 09:35:39 +0000 http://shape-of-code.coding-guidelines.com/?p=177#comment-1490 You forgot some usual assumptions: - char is 8 bits (from POSIX) - strange endianness (it seems that a lot of program uses custom function to transform data) - pointer are longer than long (IIRC since C99) The first two items are important to read binary files (e.g. images) or from network. Programs use custom function but now that all devices must share data such assumptions could slow down porting applications. You forgot some usual assumptions:
- char is 8 bits (from POSIX)
- strange endianness (it seems that a lot of program uses custom function to transform data)
- pointer are longer than long (IIRC since C99)

The first two items are important to read binary files (e.g. images) or from network. Programs use custom function but now that all devices must share data such assumptions could slow down porting applications.

]]> Comment on Designing a processor for increased source portability costs by Tony Finch http://shape-of-code.coding-guidelines.com/2010/02/designing-a-processor-for-increased-source-portability-costs/#comment-1482 Tony Finch Tue, 09 Feb 2010 08:31:50 +0000 http://shape-of-code.coding-guidelines.com/?p=177#comment-1482 x86-64 has a signed address space. http://en.wikipedia.org/wiki/X86-64#Virtual_address_space_details x86-64 has a signed address space. http://en.wikipedia.org/wiki/X86-64#Virtual_address_space_details

]]> Comment on Secret instruction sets about to make a come-back by The Shape of Code » Designing a processor for increased source portability costs http://shape-of-code.coding-guidelines.com/2010/01/secret-instruction-sets-about-to-make-a-come-back/#comment-1448 The Shape of Code » Designing a processor for increased source portability costs Tue, 09 Feb 2010 02:29:16 +0000 http://shape-of-code.coding-guidelines.com/?p=173#comment-1448 [...] vendor make it difficult for developers to port open source applications to their proprietary cpu? Keeping the instruction set secret is one technique, another is to design a cpu that breaks often relied upon assumptions that [...] [...] vendor make it difficult for developers to port open source applications to their proprietary cpu? Keeping the instruction set secret is one technique, another is to design a cpu that breaks often relied upon assumptions that [...]

]]> Comment on Secret instruction sets about to make a come-back by Derek-Jones http://shape-of-code.coding-guidelines.com/2010/01/secret-instruction-sets-about-to-make-a-come-back/#comment-1128 Derek-Jones Fri, 29 Jan 2010 12:26:34 +0000 http://shape-of-code.coding-guidelines.com/?p=173#comment-1128 I don't believe that the 'many eyes' theory of fault finding (or instruction reverse engineering) holds much water. Most people don' have the skill, inclination or time to get involved. Over time information on any secret instruction set will leak out. It only takes one skilled person with the inclination and time, or somebody to leak a confidential document. But once sufficient information is available who will write the code generator? Again given time I'm sure one or more people will write one. However, Apple's lawyers will ensure that getting hold of this compiler will be hard (assuming Apple want to keep the instruction set secret), so bugs will probably take ages to be fixed and the quality probably low. Any commercial company will not touch a 'black market' compiler and will have to strike a deal with Apple to port their application, which means Apple maintains control of the money making (the whole purpose of the exercise).. I don’t believe that the ‘many eyes’ theory of fault finding (or instruction reverse engineering) holds much water. Most people don’ have the skill, inclination or time to get involved.

Over time information on any secret instruction set will leak out. It only takes one skilled person with the inclination and time, or somebody to leak a confidential document. But once sufficient information is available who will write the code generator? Again given time I’m sure one or more people will write one. However, Apple’s lawyers will ensure that getting hold of this compiler will be hard (assuming Apple want to keep the instruction set secret), so bugs will probably take ages to be fixed and the quality probably low.

Any commercial company will not touch a ‘black market’ compiler and will have to strike a deal with Apple to port their application, which means Apple maintains control of the money making (the whole purpose of the exercise)..

]]> Comment on Minimum information needed for writing a code generator by The Shape of Code » Minimum information needed for writing a code … Zero Me http://shape-of-code.coding-guidelines.com/2010/01/minimum-information-needed-for-writing-a-code-generator/#comment-1127 The Shape of Code » Minimum information needed for writing a code … Zero Me Fri, 29 Jan 2010 08:32:12 +0000 http://shape-of-code.coding-guidelines.com/?p=175#comment-1127 [...] the original:  The Shape of Code » Minimum information needed for writing a code … tags: after-the-first, after-the-last, based-on-using, bit-thin, kettle, opposite, [...] [...] the original:  The Shape of Code » Minimum information needed for writing a code … tags: after-the-first, after-the-last, based-on-using, bit-thin, kettle, opposite, [...]

]]> Comment on Secret instruction sets about to make a come-back by Stephen Parker http://shape-of-code.coding-guidelines.com/2010/01/secret-instruction-sets-about-to-make-a-come-back/#comment-1126 Stephen Parker Fri, 29 Jan 2010 07:08:43 +0000 http://shape-of-code.coding-guidelines.com/?p=173#comment-1126 This seems unlikely. The A4 will be an ARM Cortex plus peripherals. There are too many iPhone OS developers out there for any specialised instruction set to remain secret. This seems unlikely. The A4 will be an ARM Cortex plus peripherals. There are too many iPhone OS developers out there for any specialised instruction set to remain secret.

]]> Comment on Secret instruction sets about to make a come-back by Twitter Trackbacks for The Shape of Code » Secret instruction sets about to make a come-back [coding-guidelines.com] on Topsy.com http://shape-of-code.coding-guidelines.com/2010/01/secret-instruction-sets-about-to-make-a-come-back/#comment-1123 Twitter Trackbacks for The Shape of Code » Secret instruction sets about to make a come-back [coding-guidelines.com] on Topsy.com Fri, 29 Jan 2010 05:23:40 +0000 http://shape-of-code.coding-guidelines.com/?p=173#comment-1123 [...] The Shape of Code » Secret instruction sets about to make a come-back shape-of-code.coding-guidelines.com/2010/01/secret-instruction-sets-about-to-make-a-come-back – view page – cached Home > Uncategorized > Secret instruction sets about to make a come-back [...] [...] The Shape of Code » Secret instruction sets about to make a come-back shape-of-code.coding-guidelines.com/2010/01/secret-instruction-sets-about-to-make-a-come-back – view page – cached Home > Uncategorized > Secret instruction sets about to make a come-back [...]

]]> Comment on Minimum information needed for writing a code generator by powder metallurgy « the factory | Metallurgy Material Geek http://shape-of-code.coding-guidelines.com/2010/01/minimum-information-needed-for-writing-a-code-generator/#comment-1117 powder metallurgy « the factory | Metallurgy Material Geek Fri, 29 Jan 2010 01:36:36 +0000 http://shape-of-code.coding-guidelines.com/?p=175#comment-1117 [...] The Shape of Code » Minimum information needed for writing a code … [...] [...] The Shape of Code » Minimum information needed for writing a code … [...]

]]>