Comments on: keksum, a Keccak implementation in C as standalone Unix utility: genesis http://jfxpt.com/2019/keksum-a-keccak-implementation-in-c-as-standalone-unix-utility-genesis/ The search for invariants Sun, 19 Apr 2026 03:47:16 +0000 http://polimedia.us hourly 1 By: That, in a word, is the corner that I'm painted into, even though this seems an incredibly obscure situation of having a top-level macro which inserts a top-level definition, which may or may not be previously defined and depends on the interpreter's inte http://jfxpt.com/2019/keksum-a-keccak-implementation-in-c-as-standalone-unix-utility-genesis/#comment-2989 That, in a word, is the corner that I'm painted into, even though this seems an incredibly obscure situation of having a top-level macro which inserts a top-level definition, which may or may not be previously defined and depends on the interpreter's inte Tue, 26 Nov 2024 21:54:20 +0000 http://fixpoint.welshcomputing.com/?p=70#comment-2989 [...] Result of a keksum [...] [...] Result of a keksum [...]

]]> By: File-level error recovery for keksum « Fixpoint http://jfxpt.com/2019/keksum-a-keccak-implementation-in-c-as-standalone-unix-utility-genesis/#comment-2605 File-level error recovery for keksum « Fixpoint Sat, 09 Mar 2024 21:03:30 +0000 http://fixpoint.welshcomputing.com/?p=70#comment-2605 [...] experience using and teaching with my keksum utility informed that its main deficiency was not in fact the lack of an equivalent to GNU's old [...] [...] experience using and teaching with my keksum utility informed that its main deficiency was not in fact the lack of an equivalent to GNU's old [...]

]]> By: The simplest way yet to fetch Bitcoin code « Fixpoint http://jfxpt.com/2019/keksum-a-keccak-implementation-in-c-as-standalone-unix-utility-genesis/#comment-1960 The simplest way yet to fetch Bitcoin code « Fixpoint Fri, 04 Feb 2022 00:22:04 +0000 http://fixpoint.welshcomputing.com/?p=70#comment-1960 [...] keksum - for verifying patch output hashes; as a concession to the bootstrapping conundrum, this checking is skipped with a warning displayed if this utility is not found on the system. (Signatures are always checked.) [...] [...] keksum - for verifying patch output hashes; as a concession to the bootstrapping conundrum, this checking is skipped with a warning displayed if this utility is not found on the system. (Signatures are always checked.) [...]

]]> By: EuCrypt addition: Keccak File Hashing « Ossa Sepia http://jfxpt.com/2019/keksum-a-keccak-implementation-in-c-as-standalone-unix-utility-genesis/#comment-300 EuCrypt addition: Keccak File Hashing « Ossa Sepia Wed, 08 Jul 2020 13:37:55 +0000 http://fixpoint.welshcomputing.com/?p=70#comment-300 [...] of files both raw and in the more usual numerical format (e.g. hex as given for instance by the keksum implementation). More importantly for me, Eulora's client can now check the hash of a received file [...] [...] of files both raw and in the more usual numerical format (e.g. hex as given for instance by the keksum implementation). More importantly for me, Eulora's client can now check the hash of a received file [...]

]]> By: V in Perl with parsing fix, keksum, and starter, plus the ill-fated vdiff « Fixpoint http://jfxpt.com/2019/keksum-a-keccak-implementation-in-c-as-standalone-unix-utility-genesis/#comment-230 V in Perl with parsing fix, keksum, and starter, plus the ill-fated vdiff « Fixpoint Thu, 02 Apr 2020 17:50:44 +0000 http://fixpoint.welshcomputing.com/?p=70#comment-230 [...] second patch, "v_keksum_busybox", swaps keksum and patch in for ksum and vpatch, making V presses possible again on systems with little more than [...] [...] second patch, "v_keksum_busybox", swaps keksum and patch in for ksum and vpatch, making V presses possible again on systems with little more than [...]

]]> By: Adventures in the forest of V « Fixpoint http://jfxpt.com/2019/keksum-a-keccak-implementation-in-c-as-standalone-unix-utility-genesis/#comment-222 Adventures in the forest of V « Fixpoint Tue, 31 Mar 2020 19:11:11 +0000 http://fixpoint.welshcomputing.com/?p=70#comment-222 [...] to press the Bitcoin code, or rather the VTree that grew from it, swap out the hash with my own keksum so as to avoid a hefty and otherwise unnecessary GNAT requirement, add my version of the classic [...] [...] to press the Bitcoin code, or rather the VTree that grew from it, swap out the hash with my own keksum so as to avoid a hefty and otherwise unnecessary GNAT requirement, add my version of the classic [...]

]]> By: From the forum log, 24-26 December 2019 « Fixpoint http://jfxpt.com/2019/keksum-a-keccak-implementation-in-c-as-standalone-unix-utility-genesis/#comment-171 From the forum log, 24-26 December 2019 « Fixpoint Sun, 26 Jan 2020 02:41:00 +0000 http://fixpoint.welshcomputing.com/?p=70#comment-171 [...] signed and mirrored jfw's updated keksum genesis and noted he had borked the article [...] [...] signed and mirrored jfw's updated keksum genesis and noted he had borked the article [...]

]]> By: Jacob Welsh http://jfxpt.com/2019/keksum-a-keccak-implementation-in-c-as-standalone-unix-utility-genesis/#comment-138 Jacob Welsh Thu, 19 Dec 2019 11:43:54 +0000 http://fixpoint.welshcomputing.com/?p=70#comment-138 Updated to fix the noted problems except for <a href="http://ossasepia.com/2020/04/21/ossasepia-logs-for-18-Dec-2019#1013482" rel="nofollow">improved full-program tests</a>. Updated to fix the noted problems except for improved full-program tests.

]]> By: Jacob Welsh http://jfxpt.com/2019/keksum-a-keccak-implementation-in-c-as-standalone-unix-utility-genesis/#comment-128 Jacob Welsh Sun, 15 Dec 2019 20:45:50 +0000 http://fixpoint.welshcomputing.com/?p=70#comment-128 Will Haack <a href="http://ossasepia.com/2020/04/21/ossasepia-logs-for-13-Dec-2019#1012909" rel="nofollow">reports</a> successful build and run on OSX 10.9 after removal of -static from LDFLAGS in the Makefile. Some hitches were encountered in getting the tests to run. From <a href="http://ossasepia.com/2020/04/21/ossasepia-logs-for-14-Dec-2019#1013020" rel="nofollow">the #o log</a> : <strong>diana_coman</strong>: jfw: I went through your keksum proggy and it's been quite a pleasure really; a few nitpicks on top of those: 1. why default/fallthrough on "bad option" instead of the more useful help option? 2. you have nicely \t everywhere except in usage_err in main.c where it's \s 3. just out of curiosity really since otherwise it's of course not a problem at all: since you unrolled various bits basically + made the table for rho, how/why did you choose where to stop with the unrolling anyway? and why not use the constants for iota instead of calc each time (ie lfsr)? 4. why oh why only self-gen test? not like there weren't values and/or ways to obtain them, lolz! <strong>diana_coman</strong>: jfw: anyways, once you regrind perhaps to fix the test at least, I'll gladly sign and mirror it. <strong>jfw</strong>: 1. usage_err does show the help, but returns a failure status whereas explicit help returns success. <strong>jfw</strong>: 2. mixed spaces and tab too, good catch! <strong>jfw</strong>: 3. iirc I unrolled where it would allow precomputing arithmetic and especially mod-5 ops in inner loops. Iota does relatively little work, just touching one lane per round, so I preferred to keep it in 'source' form rather than magic numbers which you'd then have to verify against someone else's. <strong>jfw</strong>: The table in rho I cribbed straight from the pdf. <strong>jfw</strong>: 4. just to clarify, I did use the provided tests for the permutation, but don't recall seeing any for the sponge, perhaps I should look harder. <strong>jfw</strong>: "It seems you forgot to update the expected test hash in test-sponge.sh so now the test will fail." - yep, that's what I meant in <a href="http://ossasepia.com/2020/04/21/ossasepia-logs-for-13-Dec-2019#1012925" rel="nofollow">http://logs.ossasepia.com/log/ossasepia/2019-12-13#1012925</a> <strong>ossabot</strong>: Logged on 2019-12-13 17:32:13 jfw: gah, one test broken anyway by the default capacity change, but otherwise working for me, let me check on those errors. <strong>jfw</strong>: "why the -Woverlength-strings in main.c, isn't that included in -pedantic" - the way the pragma works is to *disable* that warning as it's tripped by the help text. The 'wat' expands to roughly: 'why does c89 allow compilers to limit strings to something as low as 509 characters? wtf am I supposed to do about this? I have no intention of supporting such broken compilers, next they'll be limiting source file length too.' Will Haack reports successful build and run on OSX 10.9 after removal of -static from LDFLAGS in the Makefile. Some hitches were encountered in getting the tests to run.

From the #o log :

diana_coman: jfw: I went through your keksum proggy and it's been quite a pleasure really; a few nitpicks on top of those: 1. why default/fallthrough on "bad option" instead of the more useful help option? 2. you have nicely \t everywhere except in usage_err in main.c where it's \s 3. just out of curiosity really since otherwise it's of course not a problem at all: since you unrolled various bits basically + made the table for rho, how/why did you choose where to stop with the unrolling anyway? and why not use the constants for iota instead of calc each time (ie lfsr)? 4. why oh why only self-gen test? not like there weren't values and/or ways to obtain them, lolz!
diana_coman: jfw: anyways, once you regrind perhaps to fix the test at least, I'll gladly sign and mirror it.

jfw: 1. usage_err does show the help, but returns a failure status whereas explicit help returns success.
jfw: 2. mixed spaces and tab too, good catch!
jfw: 3. iirc I unrolled where it would allow precomputing arithmetic and especially mod-5 ops in inner loops. Iota does relatively little work, just touching one lane per round, so I preferred to keep it in 'source' form rather than magic numbers which you'd then have to verify against someone else's.
jfw: The table in rho I cribbed straight from the pdf.
jfw: 4. just to clarify, I did use the provided tests for the permutation, but don't recall seeing any for the sponge, perhaps I should look harder.

jfw: "It seems you forgot to update the expected test hash in test-sponge.sh so now the test will fail." - yep, that's what I meant in http://logs.ossasepia.com/log/ossasepia/2019-12-13#1012925
ossabot: Logged on 2019-12-13 17:32:13 jfw: gah, one test broken anyway by the default capacity change, but otherwise working for me, let me check on those errors.

jfw: "why the -Woverlength-strings in main.c, isn't that included in -pedantic" - the way the pragma works is to *disable* that warning as it's tripped by the help text. The 'wat' expands to roughly: 'why does c89 allow compilers to limit strings to something as low as 509 characters? wtf am I supposed to do about this? I have no intention of supporting such broken compilers, next they'll be limiting source file length too.'

]]> By: Diana Coman http://jfxpt.com/2019/keksum-a-keccak-implementation-in-c-as-standalone-unix-utility-genesis/#comment-122 Diana Coman Sat, 14 Dec 2019 11:26:05 +0000 http://fixpoint.welshcomputing.com/?p=70#comment-122 It seems you forgot to update the expected test hash in test-sponge.sh so now the test will fail. Also, why the -Woverlength-strings in main.c, isn't that included in -pedantic that I see you are using anyway in Makefile? (The "wat" comment is... very helpful, too :D ) It seems you forgot to update the expected test hash in test-sponge.sh so now the test will fail.

Also, why the -Woverlength-strings in main.c, isn't that included in -pedantic that I see you are using anyway in Makefile? (The "wat" comment is... very helpful, too :D )

]]>