Skip to contentSkip to conversation

meta-browser duplicate patch

DISCUSSION Trevor Woerner 2 participants 3 messages Jan 12, 2017

lore
Share & tools
lore
#0 lore raw

Does the do_patch logic now ignore the same patch being applied twice? I was quite sure this was a problem in the past (and if it isn't, maybe it should be?)

In commit fb4e542079958b6afde324a1e9bfdd695842aeeb
the patch recipes-browser/chromium/chromium/chromium/Do-not-depend-on-Linux-4.5.patch

and in commit c6fc9da7be64af7cd08c2bbb026e4c3ffc760687
the patch recipes-browser/chromium/chromium/chromium/unset-madv-free.patch

are exactly the same. Yet they're both happily applied without so much as a warning, not even in temp/log.do_patch.

#1 lore raw
Quoted 4 lines · Trevor Woerner

Does the do_patch logic now ignore the same patch being applied twice? I was quite sure this was a problem in the past (and if it isn't, maybe it should be?)

Quoted 8 lines

In commit fb4e542079958b6afde324a1e9bfdd695842aeeb
the patch recipes-browser/chromium/chromium/chromium/Do-not-depend-on-Linux-4.5.patch

and in commit c6fc9da7be64af7cd08c2bbb026e4c3ffc760687
the patch recipes-browser/chromium/chromium/chromium/unset-madv-free.patch

are exactly the same. Yet they're both happily applied without so much as a warning, not even in temp/log.do_patch.

Can you inspect the resulting file? Are the changes being applied twice in two *different* locations in the file?

We have a bug to fix this issue:
https://bugzilla.yoctoproject.org/show_bug.cgi?id=10450

Alex

#2 lore raw
Quoted 8 lines · Alexander Kanavin
Quoted 7 lines · Trevor Woerner

Does the do_patch logic now ignore the same patch being applied twice? I was quite sure this was a problem in the past (and if it isn't, maybe it should be?)

In commit fb4e542079958b6afde324a1e9bfdd695842aeeb
the patch recipes-browser/chromium/chromium/chromium/Do-not-depend-on-Linux-4.5.patch

Quoted 8 lines
Quoted 5 lines

and in commit c6fc9da7be64af7cd08c2bbb026e4c3ffc760687
the patch recipes-browser/chromium/chromium/chromium/unset-madv-free.patch

are exactly the same. Yet they're both happily applied without so much as a warning, not even in temp/log.do_patch.

Can you inspect the resulting file? Are the changes being applied twice in two *different* locations in the file?

Ah yes, there it is, applied twice:

	 44 #if OS(LINUX) && defined(MADV_FREE)
	 45 // Added in Linux 4.5, but it breaks the sandbox.
	 46 #undef MADV_FREE
	 47 #endif
	 48 
	 49 #if OS(LINUX) && defined(MADV_FREE)
	 50 // Added in Linux 4.5, but we don't want to depend on 4.5 at runtime
	 51 #undef MADV_FREE
	 52 #endif
Quoted 2 lines

We have a bug to fix this issue:
https://bugzilla.yoctoproject.org/show_bug.cgi?id=10450

Great, thanks :-)