I am good with this and the other changes. Thomas or anyone else have comments?
If I don't hear anything in a while, I'll just commit these to master (or Pete you can yourself.) On Tue, Feb 28, 2023 at 3:48 PM Pete Batard via Libcdio-devel < libcdio-devel@gnu.org> wrote: > The current libcdio parses Rock Ridge CE records but does not process them. > > This creates issues when trying to handle an ISO-9660 image like > Debian's 'firmware-11.6.0-amd64-netinst.iso', which can be downloaded > from > > https://cdimage.debian.org/cdimage/unofficial/non-free/cd-including-firmware/current/amd64/iso-cd/, > > since this image has some of the Rock Ridge symbolic links for the > entries found in the /firmware/ directory defined in a CE record (such > as '/firmware/firmware-ath9k-htc_1.4.0-106-gc583009+dfsg1-1_all.deb'). > > This patch series adds CE record processing and also applies some clean > up to the current Rock Ridge code. > > Note: As I found out after wasting a couple hours figuring out why one > of the CE SL entries was not processed as expected, you really don't > want to feed a char based array to from_733(), as the *signed* char > values will be sign-extended, and you may end up with a 0xffffffc4 value > instead of the expected 0x000000c4 when processing a 0xc4000000000000c4 > entry! > > Regards, > > /Pete > > Pete Batard (2): > Clean up rock.h and fix the use of signed chars in iso_su_ce_s > Add processing of Rock Ridge CE records > > include/cdio/rock.h | 18 +++++------ > lib/iso9660/rock.c | 79 ++++++++++++++++++++++++++------------------- > 2 files changed, 54 insertions(+), 43 deletions(-) > > -- > 2.39.1.windows.1 > > >
