--- Log opened Fri Sep 10 00:00:03 2021
00:32 -!- siv2r09300892 [~siv2r@103.77.37.175] has quit [Quit: The Lounge - https://thelounge.chat]
01:33 -!- dr-orlovsky [~dr-orlovs@31.14.40.18] has quit [Ping timeout: 252 seconds]
01:41 -!- dr-orlovsky [~dr-orlovs@31.14.40.18] has joined #secp256k1
02:00 -!- eli[m] [~elinixbit@2001:470:69fc:105::ba64] has quit [Quit: You have been kicked for being idle]
05:07 < elichai2> So it looks like `secp256k1_ecmult_gen_context_build` uses too much stack space for  the default value in intel SGX with some frameworks 
05:08 < elichai2> someone complained to me that libsecp is crashing for them and after a lot of gdb I found out that this was the culprit, and setting `ECMULT_GEN_PREC_BITS=2` or changing the default stack size fixed it
05:47 < sipa> another reason to make it static only :)
08:04 < gmaxwell> it's normally static, so they're already in a weird non-standard build.
08:05 < gmaxwell> a couple of the tests unfortunately use a lot of stack.
08:06 < gmaxwell> but otherwise the library should work with a very small stack.
08:35 < elichai2> gmaxwell: supposedly it's the `precj` array. and AFAIR the current default is without static ecmult precomputation
08:36 < elichai2> nope, you're right the default is with. only in rust-secp256k1 the default is without
08:48 < gmaxwell> this is part of why I want to take away these configuration options, so people stop accidentally ending up in non-standard builds.
10:48 -!- fjahr [sid374480@id-374480.highgate.irccloud.com] has quit [Ping timeout: 250 seconds]
10:52 -!- fjahr [sid374480@id-374480.highgate.irccloud.com] has joined #secp256k1
17:20 -!- belcher [~belcher@user/belcher] has quit [Ping timeout: 260 seconds]
17:34 -!- belcher [~belcher@user/belcher] has joined #secp256k1
--- Log closed Sat Sep 11 00:00:04 2021