# 00:00:46 |
sh4rm4 |
it came on the market in 1982 |
# 00:01:23 |
sh4rm4 |
and must have been relatively cheap |
# 05:57:33 |
tor |
joins #crosstool-ng |
# 06:30:33 |
kos_tom |
joins #crosstool-ng |
# 08:08:23 |
mnt_real |
quits : Quit: Ex-Chat |
# 10:02:58 |
y_morin |
joins #crosstool-ng |
# 10:16:23 |
y_morin |
is now known as: y_morin|away |
# 10:37:23 |
y_morin|away |
is now known as: y_morin |
# 11:10:54 |
y_morin |
kos_tom: ping? |
# 11:23:12 |
y_morin |
is now known as: y_morin|away |
# 11:47:36 |
kos_tom |
y_morin|away, pong! |
# 12:13:05 |
smartin |
joins #crosstool-ng |
# 12:53:47 |
y_morin|away |
is now known as: y_morin |
# 12:56:02 |
y_morin |
kos_tom: I've looked at your sh2a patches. Looks good to me. |
# 12:57:44 |
y_morin |
kos_tom: maybe some patches should be merged, or swapped, though. |
# 12:57:45 |
kos_tom |
y_morin, hmm, the sh2a symlink thing is horrible. |
# 12:58:14 |
y_morin |
kos_tom: yes, it's a kludge. But the alternative is to patch gcc. |
# 12:58:15 |
kos_tom |
y_morin, and also the generated toolchain doesn't work from a C++ point of view, and I don't know why. It's a problem related to how libgcc was built, though I don't understand the problem. |
# 12:58:30 |
kos_tom |
y_morin, the generated toolchain has some !m2a directory |
# 12:58:32 |
y_morin |
kos_tom: at least, C-only is better than nothing. |
# 12:58:36 |
kos_tom |
and I think it's related to the C++ problem. |
# 12:58:50 |
y_morin |
kos_tom: !m2a in the spec file, right? |
# 12:59:11 |
kos_tom |
!m2a is in the spec file, but there is also a directory with this name in the generated toolchain |
# 12:59:15 |
y_morin |
Note that I only looked at the patches, I did not try to build yet. |
# 12:59:31 |
kos_tom |
http://code.bulix.org/gd6bg5-80521?raw |
# 12:59:46 |
y_morin |
kos_tom: so if we patch gcc so that the spec file does not have !m2a, we can get rid of the symlink. |
# 13:00:07 |
kos_tom |
the C++ stuff (libsupc++, libstdc++) is in !m2a |
# 13:00:42 |
y_morin |
Oh, a directory named '!m2a' Arg. |
# 13:02:04 |
y_morin |
kos_tom: also, it is static-only. Can't use shared? |
# 13:05:57 |
kos_tom |
y_morin, no, the FDPIC support for this architecture is not in uClibc mainline |
# 13:06:18 |
kos_tom |
y_morin, CodeSourcery has written years ago some patches, but there are not in mainline uClibc, and for the moment, I was only interested in getting static lib working. |
# 13:06:27 |
kos_tom |
remember it's a no-MMU platform, so shared libraries are complicated :) |
# 13:06:37 |
y_morin |
OK. But then, you have: [*] Build shared libraries |
# 13:07:16 |
kos_tom |
hm, maybe my sample config is not completely up-to-date |
# 13:07:32 |
kos_tom |
my .config has # CT_SHARED_LIBS is not set |
# 13:08:05 |
kos_tom |
ah, yes, my sample in my working tree is correct (SHARED_LIBS disabled), but not in the patch I gave you |
# 13:08:23 |
kos_tom |
my Mercurial workflow is completely random... |
# 13:08:34 |
y_morin |
kos_tom: no problem so far! I'm just trying to see discrpancy amopgst config options, and what we get in the end... |
# 13:09:42 |
kos_tom |
yep, ok. The output from above is obviously with CT_SHARED_LIBS=n |
# 14:01:41 |
y_morin |
is now known as: y_morin|away |
# 15:04:48 |
y_morin|away |
is now known as: y_morin |
# 15:16:04 |
y_morin |
is now known as: y_morin|away |
# 15:57:38 |
smartin |
quits : Ping timeout: 252 seconds |
# 16:17:54 |
mnt_real |
joins #crosstool-ng |
# 16:25:08 |
kos_tom |
quits : Quit: Leaving |
# 18:29:07 |
y_morin|away |
is now known as: y_morin |
# 18:40:03 |
y_morin |
is now known as: y_morin|away |
# 18:59:27 |
kos_tom |
joins #crosstool-ng |
# 20:35:52 |
tor |
quits : Quit: Ex-Chat |
# 21:12:43 |
y_morin|away |
is now known as: y_morin |
# 21:36:02 |
y_morin |
is now known as: y_morin|away |
# 21:36:07 |
y_morin|away |
is now known as: y_morin |
# 22:12:57 |
kos_tom |
quits : Ping timeout: 255 seconds |
# 23:53:05 |
y_morin |
is now known as: y_morin|away |
# 23:57:40 |
y_morin|away |
is now known as: y_morin |