--- Log opened Mon Nov 25 00:00:18 2019
00:01 -!- jonatack_ [~jon@37.170.7.180] has joined #secp256k1
00:04 -!- jonatack [~jon@37.164.37.175] has quit [Ping timeout: 265 seconds]
01:29 < midnightmagic> \o/
02:19 < gmaxwell> https://twitter.com/travisci/status/1194332270301040640  < S390x linux, IIRC the stack is a linked list, lol.. might be fun to test on.
02:19 < gmaxwell> (I've actually tested libsecp256k1 on s390, but years ago.
02:19 < gmaxwell> )
03:39 < elichai2> gmaxwell: https://github.com/bitcoin/bitcoin/pull/17402 :)
03:39 < elichai2> but that's powerpc, not s390x
03:39 < elichai2> is s390x LE?
03:39 < elichai2> *BE
03:39 < gmaxwell> IIRC it's BE.
03:40 < elichai2> that's good. we looked for BE testing in core
03:44 -!- cfields [~cfields@unaffiliated/cfields] has quit [Remote host closed the connection]
03:48 -!- cfields_ [~cfields@unaffiliated/cfields] has joined #secp256k1
04:09 < elichai2> gmaxwell: yep :) https://travis-ci.org/elichai/Test-travis-BE/builds/616645984#L11 i'll PR to bitcoin now :)
04:36 < gmaxwell> I don't recall if there is anything else particularly weird about s390x. IIRC it's an unsigned char, LP64, BE. At least if the host OS is fedora/centos I expect bitcoin will just work.
06:02 < elichai2> I'm experimenting with ECMH (https://github.com/bitcoin-core/secp256k1/pull/477) and I see that he's always normalizing everything, which sucks. so I'm thinking of replacing it with an opaque type to represent `secp256k1_gej`. the problem is that it has different size per platform.  so i'm thinking of making it the size of a 64bit machine and putting some if/defs when converting from/to gej.
06:02 < elichai2> wanted to hear if anyone has any thoughts :) (also thinking of maybe doing some sort of batch addition, not sure if it's worth it though)
06:03 < sipa> elichai2: use secp256k1_fe_storage to store x, y, z directly
06:04 < sipa> which is guaranteed to be 32 bytes per element
06:04 < elichai2> doesn't this require it to be normalized?
06:05 < sipa> oh, yes
06:05 < sipa> don't worry about that
06:05 < sipa> normalization is trivial
06:05 < sipa> it's something to avoid where possible on the scale of individual fe_add operations etc
06:05 < sipa> but not at the level of a point addition
06:06 < elichai2> oh really? why is it less important in point addition?
06:06 < elichai2> hmm just seeing that point addition starts with a `secp256k1_fe_normalize_weak`
06:07 < elichai2> so I could also just add a variant without those because i'm taking from storage  (and add a VERIFY check...)
06:08 < sipa> normalization is negligable compared to point addition
06:27 -!- sipa [~pw@gateway/tor-sasl/sipa1024] has quit [Ping timeout: 260 seconds]
07:29 < elichai2> I actually think most of the work there is in the sqrt
07:31 < elichai2> `field_sqrt: min 4.10us / avg 4.22us / max 4.42us` `group_add_var: min 0.286us / avg 0.290us / max 0.295us`
08:47 -!- sipa [~pw@gateway/tor-sasl/sipa1024] has joined #secp256k1
09:41 -!- jonatack_ [~jon@37.170.7.180] has quit [Quit: jonatack_]
09:42 -!- jonatack [~jon@37.170.7.180] has joined #secp256k1
12:01 -!- jonatack_ [~jon@37.173.132.239] has joined #secp256k1
12:05 -!- jonatack [~jon@37.170.7.180] has quit [Ping timeout: 276 seconds]
12:46 -!- belcher [~belcher@unaffiliated/belcher] has quit [Quit: Leaving]
15:27 -!- waxwing [~waxwing@unaffiliated/waxwing] has quit [Quit: ZNC 1.6.5 - http://znc.in]
15:27 -!- waxwing [~waxwing@193.29.57.116] has joined #secp256k1
15:29 -!- waxwing [~waxwing@193.29.57.116] has quit [Changing host]
15:29 -!- waxwing [~waxwing@unaffiliated/waxwing] has joined #secp256k1
15:46 -!- belcher [~belcher@unaffiliated/belcher] has joined #secp256k1
19:26 -!- reallll [~belcher@unaffiliated/belcher] has joined #secp256k1
19:30 -!- belcher [~belcher@unaffiliated/belcher] has quit [Ping timeout: 265 seconds]
19:57 -!- jonatack_ [~jon@37.173.132.239] has quit [Read error: Connection reset by peer]
20:49 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Ping timeout: 246 seconds]
21:22 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #secp256k1
22:27 -!- jonatack_ [~jon@37.173.132.239] has joined #secp256k1
22:28 -!- jonatack_ [~jon@37.173.132.239] has quit [Client Quit]
22:28 -!- jonatack [~jon@37.173.132.239] has joined #secp256k1
--- Log closed Tue Nov 26 00:00:19 2019