-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathvoid-mklive:pull:128.html
178 lines (178 loc) · 7.99 KB
/
void-mklive:pull:128.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
<!DOCTYPE html>
<html xmlns="http://www.w3.org/1999/xhtml" lang="en">
<head>
<meta charset="utf-8" />
<title>Re: [voidlinux/void-mklive] Cleanup Phase I (#128)</title>
<style>section{white-space:pre-wrap;}</style>
</head>
<body>
<main>
<h1>Re: [voidlinux/void-mklive] Cleanup Phase I (#128)</h1>
<section id="post1">
<h5>Michael Aldridge at <a href="#post1">Mon, 21 Aug 2017 02:52:27 +0000 (UTC)</a></h5>
@the-maldridge requested your review on: voidlinux/void-mklive#128 Cleanup Phase I.
</section><section id="post2">
<h5>Michael Aldridge at <a href="#post2">Sun, 20 Aug 2017 21:10:20 -0700</a></h5>
@the-maldridge pushed 1 commit.
5791d18 Makefile: add option to set mirror across build
</section><section id="post3">
<h5>Enno Boland at <a href="#post3">Sun, 20 Aug 2017 22:46:25 -0700</a></h5>
Gottox approved this pull request.
Looks good so far =F0=9F=91=8D Added a few further ideas though.
> + # prepared. This function takes 2 arguments, the location to
+ # chroot to and the command to run.
+
+ # This is an idempotent function, it is safe to call every time
+ # before entering the chroot. This has the advantage of making
+ # execution in the chroot appear as though it "Just Works(tm)".
+ register_binfmt
+
+ # Before we step into the chroot we need to make sure the
+ # pseudo-filesystems are ready to go. Not all commands will need
+ # this, but its still a good idea to call it here anyway.
+ mount_pseudofs
+
+ # With assurance that things will run now we can jump into the
+ # chroot and run stuff!
+ chroot "$1" sh -c "$2"
I had the idea to use xbps-uchroot instead of chroot, this unshares the m=
ountspace and bindmounts /dev /prov /sys. Unfortunately iirc there were s=
ome issues with it that needed to be fixed in uchroot.
> @@ -86,9 +81,10 @@ _EOF
exit 0
}
=
-#
-# main()
-#
+# ########################################
+# SCRIPT EXECUTION STARTS HERE
+# ########################################
+
while getopts "b:B:o:r:s:hV" opt; do
Maybe we can use the getopt/getopts combination we use in xbps-src here t=
oo.
-- =
You are receiving this because your review was requested.
Reply to this email directly or view it on GitHub:
https://github.com/voidlinux/void-mklive/pull/128#pullrequestreview-57406=
026=
</section><section id="post4">
<h5>Michael Aldridge at <a href="#post4">Mon, 21 Aug 2017 05:47:40 +0000 (UTC)</a></h5>
the-maldridge commented on this pull request.
> + # prepared. This function takes 2 arguments, the location to
+ # chroot to and the command to run.
+
+ # This is an idempotent function, it is safe to call every time
+ # before entering the chroot. This has the advantage of making
+ # execution in the chroot appear as though it "Just Works(tm)".
+ register_binfmt
+
+ # Before we step into the chroot we need to make sure the
+ # pseudo-filesystems are ready to go. Not all commands will need
+ # this, but its still a good idea to call it here anyway.
+ mount_pseudofs
+
+ # With assurance that things will run now we can jump into the
+ # chroot and run stuff!
+ chroot "$1" sh -c "$2"
Does xbps-uchroot play nice with qemu-user-static?
</section><section id="post5">
<h5>Enno Boland at <a href="#post5">Mon, 21 Aug 2017 06:15:22 +0000 (UTC)</a></h5>
Gottox commented on this pull request.
> + # prepared. This function takes 2 arguments, the location to
+ # chroot to and the command to run.
+
+ # This is an idempotent function, it is safe to call every time
+ # before entering the chroot. This has the advantage of making
+ # execution in the chroot appear as though it "Just Works(tm)".
+ register_binfmt
+
+ # Before we step into the chroot we need to make sure the
+ # pseudo-filesystems are ready to go. Not all commands will need
+ # this, but its still a good idea to call it here anyway.
+ mount_pseudofs
+
+ # With assurance that things will run now we can jump into the
+ # chroot and run stuff!
+ chroot "$1" sh -c "$2"
xbps-uchroot is actually chroot + unshare. I don't know if there are implications to qemu
</section><section id="post6">
<h5>Michael Aldridge at <a href="#post6">Sun, 20 Aug 2017 23:40:49 -0700</a></h5>
the-maldridge commented on this pull request.
> @@ -86,9 +81,10 @@ _EOF
exit 0
}
-#
-# main()
-#
+# ########################################
+# SCRIPT EXECUTION STARTS HERE
+# ########################################
+
while getopts "b:B:o:r:s:hV" opt; do
I looked into this and it seems that the important options are being mirrored across. I don't think copying over more options that are already copied over buys us anything. I certainly don't think the dance that xbps-src does with preprocessing is worth it here.
</section><section id="post7">
<h5>Michael Aldridge at <a href="#post7">Mon, 21 Aug 2017 07:04:25 +0000 (UTC)</a></h5>
@the-maldridge pushed 1 commit.
4f514f5 Explicitly set number of xz compressor threads
</section><section id="post8">
<h5>Enno Boland at <a href="#post8">Mon, 21 Aug 2017 08:59:50 +0000 (UTC)</a></h5>
Gottox commented on this pull request.
> @@ -23,6 +23,7 @@ ALL_CLOUD_IMAGES=$(foreach cloud,$(CLOUD_IMGS),void-$(cloud)-$(DATE).tar.gz)
SUDO := sudo
XBPS_REPOSITORY := -r https://lug.utdallas.edu/mirror/void/current -r https://lug.utdallas.edu/mirror/void/current/musl -r https://lug.utdallas.edu/mirror/void/current/aarch64
+COMPRESSOR_THREADS=2
Can we dynamicly use the number of CPUs in the system? Does this make sense?
And I'm curious: Does running multithreaded xz result always result in an identical file with the same checksum?
</section><section id="post9">
<h5>Michael Aldridge at <a href="#post9">Mon, 21 Aug 2017 20:01:12 -0700</a></h5>
the-maldridge commented on this pull request.
> @@ -23,6 +23,7 @@ ALL_CLOUD_IMAGES=$(foreach cloud,$(CLOUD_IMGS),void-$(cloud)-$(DATE).tar.gz)
SUDO := sudo
XBPS_REPOSITORY := -r https://lug.utdallas.edu/mirror/void/current -r https://lug.utdallas.edu/mirror/void/current/musl -r https://lug.utdallas.edu/mirror/void/current/aarch64
+COMPRESSOR_THREADS=2
setting this to 0 will dynamically use the maximum number of threads. The problem with this is that it also takes a lot of memory to do this.
</section><section id="post10">
<h5>Enno Boland at <a href="#post10">Tue, 22 Aug 2017 06:14:40 +0000 (UTC)</a></h5>
Gottox commented on this pull request.
> @@ -23,6 +23,7 @@ ALL_CLOUD_IMAGES=$(foreach cloud,$(CLOUD_IMGS),void-$(cloud)-$(DATE).tar.gz)
SUDO := sudo
XBPS_REPOSITORY := -r https://lug.utdallas.edu/mirror/void/current -r https://lug.utdallas.edu/mirror/void/current/musl -r https://lug.utdallas.edu/mirror/void/current/aarch64
+COMPRESSOR_THREADS=2
Ok.
</section><section id="post11">
<h5>Enno Boland at <a href="#post11">Mon, 21 Aug 2017 23:15:23 -0700</a></h5>
Gottox commented on this pull request.
> + # prepared. This function takes 2 arguments, the location to
+ # chroot to and the command to run.
+
+ # This is an idempotent function, it is safe to call every time
+ # before entering the chroot. This has the advantage of making
+ # execution in the chroot appear as though it "Just Works(tm)".
+ register_binfmt
+
+ # Before we step into the chroot we need to make sure the
+ # pseudo-filesystems are ready to go. Not all commands will need
+ # this, but its still a good idea to call it here anyway.
+ mount_pseudofs
+
+ # With assurance that things will run now we can jump into the
+ # chroot and run stuff!
+ chroot "$1" sh -c "$2"
I think we should keep that change in our mind, But it shouldn't be part of this PR.
</section><section id="post12">
<h5>Enno Boland at <a href="#post12">Mon, 21 Aug 2017 23:18:05 -0700</a></h5>
Ok to merge for me.
</section><section id="post13">
<h5>Michael Aldridge at <a href="#post13">Tue, 22 Aug 2017 22:11:46 -0700</a></h5>
@the-maldridge pushed 1 commit.
6335173 README.md: update minimally
</section><section id="post14">
<h5>Michael Aldridge at <a href="#post14">Wed, 23 Aug 2017 05:11:58 +0000 (UTC)</a></h5>
Merged #128.
</section>
</main>
<nav><a href="index.html">Issues list</a></nav>
</body>
</html>