1 |
Dale wrote: |
2 |
> Jack wrote: |
3 |
>> On 2/28/21 4:53 PM, Dale wrote: |
4 |
>>> Howdy, |
5 |
>>> |
6 |
>>> For the past couple weeks, this package has failed to update because it |
7 |
>>> is the wrong size. This is from the logs for today. |
8 |
>>> |
9 |
>>> |
10 |
>>> 2021-02-22 07:20:45 (96.2 KB/s) - |
11 |
>>> ‘/var/cache/portage/distfiles/asciidoc-9.0.5.tar.gz.__download__’ saved |
12 |
>>> [1119288] |
13 |
>>> |
14 |
>>> !!! Fetched file: asciidoc-9.0.5.tar.gz VERIFY FAILED! |
15 |
>>> !!! Reason: Filesize does not match recorded size |
16 |
>>> !!! Got: 1119288 |
17 |
>>> !!! Expected: 1119318 |
18 |
>>> Refetching... File renamed to |
19 |
>>> '/var/cache/portage/distfiles/asciidoc-9.0.5.tar.gz._checksum_failure_.ngu5ov57' |
20 |
>>> |
21 |
>>> |
22 |
>>> !!! Couldn't download 'asciidoc-9.0.5.tar.gz'. Aborting. |
23 |
>>> |
24 |
>>> |
25 |
>>> I found a bug that I think addresses this issue. It appears it should |
26 |
>>> be fixed but I've synced twice since then and it still fails. Here's |
27 |
>>> the bug. |
28 |
>>> |
29 |
>>> https://bugs.gentoo.org/770841 |
30 |
>>> |
31 |
>>> I've also tried different versions with no change. Any one else running |
32 |
>>> into this? Is something wrong on my end that is preventing the fix from |
33 |
>>> reaching me? Am I missing something else? |
34 |
>>> |
35 |
>>> Thanks. |
36 |
>>> |
37 |
>>> Dale |
38 |
>> I just installed asciidoc-9.0.4 and then 9.0.5 with no problems. |
39 |
>> However, something seems odd, since those ebuilds are dated 2/14, so |
40 |
>> not recently updated. That bug looks like they identified the |
41 |
>> necessary fixes, but didn't actually apply them yet. I think what |
42 |
>> happened is that upstream made those changes talked about in the bug, |
43 |
>> but the new tarballs have not hit all the mirrors yet. If you happen |
44 |
>> to hit a mirror with the old files, the emerge will work OK. If you |
45 |
>> hit a mirror with the new tarballs, you'll get the failure you got. |
46 |
>> Since you already have the new tarball, maybe you just need to |
47 |
>> manually edit the ebuild per the bug, and rerun "ebuild |
48 |
>> path/to/file.ebuild manifest" to manually update the hash for the |
49 |
>> tarball, if you trust the tarball you have. |
50 |
>> |
51 |
>> Jack |
52 |
>> |
53 |
>> |
54 |
>> |
55 |
> |
56 |
> That makes sense. It explains why the bug says it is fixed but I'm not |
57 |
> getting the results. So, I did the manifest thing to force it to build |
58 |
> and install anyway. Now I get this: |
59 |
> |
60 |
> |
61 |
>>>> Verifying ebuild manifests |
62 |
>>>> Emerging (1 of 1) app-text/asciidoc-9.0.5::gentoo |
63 |
>>>> Failed to emerge app-text/asciidoc-9.0.5, Log file: |
64 |
>>>> '/var/log/portage/app-text:asciidoc-9.0.5:20210301-003627.log' |
65 |
>>>> Jobs: 0 of 1 complete, 1 failed Load avg: 10.9, |
66 |
> 10.6, 10.7 |
67 |
> * Package: app-text/asciidoc-9.0.5 |
68 |
> * Repository: gentoo |
69 |
> * Maintainer: marcec@×××.de proxy-maint@g.o |
70 |
> * USE: abi_x86_64 amd64 elibc_glibc kernel_linux |
71 |
> python_single_target_python3_8 userland_GNU |
72 |
> * FEATURES: network-sandbox preserve-libs sandbox userpriv usersandbox |
73 |
> * Using python3.8 to build |
74 |
>>>> Unpacking source... |
75 |
>>>> Unpacking asciidoc-9.0.5.tar.gz to |
76 |
> /var/tmp/portage/app-text/asciidoc-9.0.5/work |
77 |
>>>> Source unpacked in /var/tmp/portage/app-text/asciidoc-9.0.5/work |
78 |
> * ERROR: app-text/asciidoc-9.0.5::gentoo failed (prepare phase): |
79 |
> * The source directory |
80 |
> '/var/tmp/portage/app-text/asciidoc-9.0.5/work/asciidoc-py3-9.0.5' |
81 |
> doesn't exist |
82 |
> * |
83 |
> * Call stack: |
84 |
> * ebuild.sh, line 762: Called __ebuild_main 'prepare' |
85 |
> * phase-functions.sh, line 1050: Called __dyn_prepare |
86 |
> * phase-functions.sh, line 384: Called die |
87 |
> * The specific snippet of code: |
88 |
> * die "The source directory '${S}' doesn't exist" |
89 |
> * |
90 |
> * If you need support, post the output of `emerge --info |
91 |
> '=app-text/asciidoc-9.0.5::gentoo'`, |
92 |
> * the complete build log and the output of `emerge -pqv |
93 |
> '=app-text/asciidoc-9.0.5::gentoo'`. |
94 |
> * The complete build log is located at |
95 |
> '/var/log/portage/app-text:asciidoc-9.0.5:20210301-003627.log'. |
96 |
> * For convenience, a symlink to the build log is located at |
97 |
> '/var/tmp/portage/app-text/asciidoc-9.0.5/temp/build.log'. |
98 |
> * The ebuild environment file is located at |
99 |
> '/var/tmp/portage/app-text/asciidoc-9.0.5/temp/environment'. |
100 |
> * Working directory: '/var/tmp/portage/app-text/asciidoc-9.0.5/homedir' |
101 |
> * S: '/var/tmp/portage/app-text/asciidoc-9.0.5/work/asciidoc-py3-9.0.5' |
102 |
> * |
103 |
> * The following package has failed to build, install, or execute postinst: |
104 |
> * |
105 |
> * (app-text/asciidoc-9.0.5:0/0::gentoo, ebuild scheduled for merge), |
106 |
> Log file: |
107 |
> * '/var/log/portage/app-text:asciidoc-9.0.5:20210301-003627.log' |
108 |
> * |
109 |
> root@fireball / # |
110 |
> |
111 |
> |
112 |
> It seems the ebuild or the tarball I'm getting isn't right. By the way, |
113 |
> when I look, that directory is there. I'm not sure why it says it |
114 |
> doesn't exist. It seems this is a much bigger problem or I need to sync |
115 |
> and get tarballs from somewhere else. Something fishy somewhere. |
116 |
> |
117 |
> Thanks for the help. Will try changing server sources next. |
118 |
> |
119 |
> Dale |
120 |
> |
121 |
> :-) :-) |
122 |
> |
123 |
|
124 |
|
125 |
I tried syncing with different servers, I think some are bad out of |
126 |
date, but it hasn't helped. I either run into a failed verify or the |
127 |
error about a directory not existing, that does exist by the way. If |
128 |
someone else runs into this, my temporary solution is to stick with |
129 |
asciidoc-9.0.2 and mask the rest. Maybe when it updates to 9.0.6 or |
130 |
something, it will be fixed. |
131 |
|
132 |
I'm not sure what is going on with the servers but when I switched to |
133 |
one of them, recent updates didn't even exist. I just did a KDE and I |
134 |
think Firefox and neither seemed to be in the tree from those servers. |
135 |
Thing is, I didn't note which server it was. If someone notices that |
136 |
their syncing isn't working right, may want to switch and find one that |
137 |
is up to date. There's something fishy going on and I don't know what |
138 |
to report. I might add, I found a bad mirror not long ago and it was |
139 |
removed. It is the US based servers. I didn't try the other |
140 |
countries. If someone has a way to test them, I used mirrorselect -r -i |
141 |
to select a server. |
142 |
|
143 |
Thanks Jack for the reply. It helped to get info from someone else. |
144 |
|
145 |
Dale |
146 |
|
147 |
:-) :-) |