[OE-core] [PATCH] [PATCH] wpa-supplicant: fix the bug for PATCHTOOL = "patch"

[Peter Kjellerstedt]

> -----Original Message-----
> From: Richard Purdie [mailto:richard.purdie at linuxfoundation.org]
> Sent: den 5 juni 2018 12:50
> To: Peter Kjellerstedt <peter.kjellerstedt at axis.com>; Hong Liu
> <hongl.fnst at cn.fujitsu.com>; openembedded-core at lists.openembedded.org
> Subject: Re: [OE-core] [PATCH] [PATCH] wpa-supplicant: fix the bug for
> PATCHTOOL = "patch"
> 
> On Tue, 2018-06-05 at 10:43 +0000, Peter Kjellerstedt wrote:
> > > -----Original Message-----
> > > From: openembedded-core-bounces at lists.openembedded.org
> > > [mailto:openembedded-core-bounces at lists.openembedded.org] On Behalf
> > > Of
> > > Hong Liu
> > > Sent: den 5 juni 2018 10:11
> > > To: openembedded-core at lists.openembedded.org
> > > Subject: [OE-core] [PATCH] [PATCH] wpa-supplicant: fix the bug for
> > > PATCHTOOL = "patch"
> > >
> > > When switch PATCHTOOL to patch, applying 'key-replay-cve-
> > > multiple.patch' failed:
> > >
> > > checking file src/ap/ieee802_11.c
> > > checking file src/ap/wpa_auth.c
> > > checking file src/ap/wpa_auth.h
> > > checking file src/ap/wpa_auth_ft.c
> > > checking file src/ap/wpa_auth_i.h
> > > checking file src/common/wpa_common.h
> > > checking file src/rsn_supp/wpa.c
> > > checking file src/rsn_supp/wpa_i.h
> > > checking file src/rsn_supp/wpa.c
> > > Hunk #1 FAILED at 709.
> > > Hunk #2 FAILED at 757.
> > > Hunk #3 succeeded at 840 (offset -12 lines).
> > > Hunk #4 FAILED at 868.
> > > Hunk #5 FAILED at 900.
> > > Hunk #6 FAILED at 924.
> > > Hunk #7 succeeded at 1536 (offset -38 lines).
> > > Hunk #8 FAILED at 2386.
> > > Hunk #9 FAILED at 2920.
> > > Hunk #10 succeeded at 2940 (offset -46 lines).
> > > Hunk #11 FAILED at 2998.
> > > 8 out of 11 hunks FAILED
> > > checking file src/rsn_supp/wpa_i.h
> > > Hunk #1 FAILED at 32.
> > > 1 out of 1 hunk FAILED
> > > checking file src/common/wpa_common.h
> > > Hunk #1 succeeded at 215 with fuzz 1.
> > > checking file src/rsn_supp/wpa.c
> > > checking file src/rsn_supp/wpa_i.h
> > > checking file src/ap/wpa_auth.c
> > > Hunk #1 succeeded at 1898 (offset -3 lines).
> > > Hunk #2 succeeded at 2470 (offset -3 lines).
> > > checking file src/rsn_supp/tdls.c
> > > checking file wpa_supplicant/wnm_sta.c
> > > checking file src/rsn_supp/wpa.c
> > > Hunk #1 succeeded at 2378 (offset -62 lines).
> > > checking file src/rsn_supp/wpa_ft.c
> > > checking file src/rsn_supp/wpa_i.h
> > > Hunk #1 succeeded at 123 (offset -5 lines).
> > >
> > > So split the wpa-supplicant/key-replay-cve-multiple to 8 patches.
> >
> > Why does it need to be split into eight separate patches? Isn't it
> > just a case of having to regenerate the patch so that the hunk
> > contexts match the current code?
> 
> You're technically right but I think separate patches may be a lot
> clearer...
> 
> Cheers,
> 
> Richard

I should probably have looked at the patches and not just the new 
file names. I just assumed the original patch was split along the 
file borders, but I now see that it was actually split according to 
functionality. However, it might be an idea to let git format-patch 
regenerate the file names to better indicate their contents.

//Peter