| 30 May 2022 |
| Magic_RB joined the room. | 22:09:38 |
| 31 May 2022 |
| fufexan joined the room. | 10:20:46 |
fufexan | my board came today 😄 | 10:21:16 |
fufexan | sadly can't test it as I don't have a HDMI monitor lying around, nor HDMI->VGA adapter. also no USB-C multiplexer (if it can even output over usb-c).
I'll look into it more when I get hold of one of those things, or after my exam session ends (late June) | 10:22:37 |
j-k | you can use ssh
https://rvspace.org/en/Product/VisionFive/Technical_Documents/VisionFive_Single_Board_Computer_Quick_Start_Guide#using-ssh-over-ethernet
or usb to serial
https://rvspace.org/en/Product/VisionFive/Technical_Documents/VisionFive_Single_Board_Computer_Quick_Start_Guide#using-a-usb-to-serial-converter
for a terminal session | 11:00:59 |
matthewcroughan - nix.how | In reply to @astro:envs.net mine arrived today and a few fellow hackers immediately got interested when I unpacked it hehehe | 19:54:10 |
matthewcroughan - nix.how | nice | 19:54:13 |
matthewcroughan - nix.how | https://youtu.be/wPrUmViN_5c | 19:54:49 |
matthewcroughan - nix.how | "RISC is gonna change everything" | 19:54:54 |
ius | In reply to @yuka:yuka.dev maybe a dumb question, did you use thermal paste for the heatsink? Another dumb one, assuming this is about VisionFive: the heatsink? Are they shipping the board with a heatsink now? | 20:08:01 |
ius | j-k: Sorry, catching up a month worth of backlog - noticed you had issues building with binfmt - did you patch qemu to ignore the threading assertion? That's the first thing I ran into, apart from a ~1-3 other bugs (of which 1 is fixed upstream)
diff --git a/plugins/core.c b/plugins/core.c
index 6b2490f973..341eefdd6b 100644
--- a/plugins/core.c
+++ b/plugins/core.c
@@ -219,7 +219,6 @@ void qemu_plugin_vcpu_init_hook(CPUState *cpu)
plugin_cpu_update__locked(&cpu->cpu_index, NULL, NULL);
success = g_hash_table_insert(plugin.cpu_ht, &cpu->cpu_index,
&cpu->cpu_index);
- g_assert(success);
qemu_rec_mutex_unlock(&plugin.lock);
plugin_vcpu_cb__simple(cpu, QEMU_PLUGIN_EV_VCPU_INIT);
| 20:11:11 |
ius | * j-k: Sorry, catching up a month worth of backlog - noticed you had issues building with binfmt - did you patch qemu to ignore the vcpu assertion? That's the first thing I ran into, apart from a ~1-3 other bugs (of which 1 is fixed upstream)
diff --git a/plugins/core.c b/plugins/core.c
index 6b2490f973..341eefdd6b 100644
--- a/plugins/core.c
+++ b/plugins/core.c
@@ -219,7 +219,6 @@ void qemu_plugin_vcpu_init_hook(CPUState *cpu)
plugin_cpu_update__locked(&cpu->cpu_index, NULL, NULL);
success = g_hash_table_insert(plugin.cpu_ht, &cpu->cpu_index,
&cpu->cpu_index);
- g_assert(success);
qemu_rec_mutex_unlock(&plugin.lock);
plugin_vcpu_cb__simple(cpu, QEMU_PLUGIN_EV_VCPU_INIT);
| 20:11:30 |
ius | Also: rebased my branch about ~two months forward, which resulting in exciting new failures (probably qemu bugs): e.g. a free() related test of coreutils which consistently fails - let me know if someone has spent any effort debugging it. | 20:13:30 |
ius | * Also: rebased my branch about ~two months forward, which results in exciting new failures (probably qemu bugs): e.g. a free() related test of coreutils which consistently fails - let me know if someone has spent any effort debugging it. | 20:13:48 |
| 1 Jun 2022 |
j-k | In reply to @ius:nltrix.net
j-k: Sorry, catching up a month worth of backlog - noticed you had issues building with binfmt - did you patch qemu to ignore the vcpu assertion? That's the first thing I ran into, apart from a ~1-3 other bugs (of which 1 is fixed upstream)
diff --git a/plugins/core.c b/plugins/core.c
index 6b2490f973..341eefdd6b 100644
--- a/plugins/core.c
+++ b/plugins/core.c
@@ -219,7 +219,6 @@ void qemu_plugin_vcpu_init_hook(CPUState *cpu)
plugin_cpu_update__locked(&cpu->cpu_index, NULL, NULL);
success = g_hash_table_insert(plugin.cpu_ht, &cpu->cpu_index,
&cpu->cpu_index);
- g_assert(success);
qemu_rec_mutex_unlock(&plugin.lock);
plugin_vcpu_cb__simple(cpu, QEMU_PLUGIN_EV_VCPU_INIT);
hmm yeah I doubt I'm using that patch . I'll have a play with that | 08:21:17 |
| tinybronca joined the room. | 09:06:20 |
| 5 Jun 2022 |
matthewcroughan - nix.how | j-k: Zhaofeng Li did you know that nix build nixpkgs#pkgsCross.riscv64.pkgsStatic.nix works? | 12:24:46 |
matthewcroughan - nix.how | not only cross compiling, but statically compiling too | 12:25:01 |
matthewcroughan - nix.how | * not only cross compiling, but statically linking too | 12:32:38 |
matthewcroughan - nix.how | I just the result to fufexan, he ran it on Fedora on the VisionFive board, it worked. | 12:35:44 |
fufexan | should've mentioned, I haven't run it on the board yet | 12:36:08 |
fufexan | I'm leaving for another city soon so it's packed rn | 12:36:36 |
matthewcroughan - nix.how | ah okay, so the screenshot you sent was through binfmt on an x86 box? | 12:36:38 |
fufexan | yes | 12:36:41 |
matthewcroughan - nix.how | If it doesn't work when it gets there then I'll be very surprised :P | 12:36:51 |
fufexan | I'll run it real quick to test | 12:36:51 |
fufexan |  Download image.png | 12:40:25 |
fufexan | sure enough, it works | 12:40:32 |
matthewcroughan - nix.how | The blog post I write this month, is going to be amazing. | 12:44:30 |
matthewcroughan - nix.how | Statically linking and cross compiling Nix itself for usage on any distribution for the VisionFive board is so simple, I'm just blown away. | 12:44:59 |