extracting milions of symlinks

memory can be freed also during extraction in case it is not needed anymore? Thanks for your help! Best regards, Ondrej Dubaj

[GNU tar-1.32] testsuite 174 failed on i686 architecture

Hi, I came across a problem during tar testsuite +++ /builddir/build/BUILD/tar-1.32/tests/testsuite.dir/at-groups/174/stdout 2020-02-04 09:02:50.834577607 + @@ -3,6 +3,6 @@ baz comparing foo -foo: Size differs +foo: Contents differ baz 174. sptrdiff01.at:26: 174. file truncated in da

Disk read error interpreted as "File shrank" + not saying it is "Padding with zeros", but padds silently

and creating an archive, then it prints "file shrank" - then it writes zeros (aka padding) according to initial file size (but does not print that message) - This happens in most cases (due to tar-block-size / disk-block-size / read-shift-by-512-bytes interaction) - I provided a reproducer which shows under which circumstances it correctly prints "Read error at byte…" Expected results: - When there is a read error, THEN tar shall report a read error - When there is a read error, THEN tar shall NOT report a "file shrank" - In addition it SHALL print "Padding with zeros". This is missing currently. Regards, Ondrej Dubaj

tar attempts to drop privileges for remote command but mishandles groups

order of the setuid() and setgid() -- since sys_reset_uid_gid does setuid first, by the time it runs setgid, the process no longer has permissions to drop the saved set-group-ID. The reproducer above does not demonstrate this second problem specifically, let's just say the order should be reversed. Regards, Ondrej Dubaj

Re: tar attempts to drop privileges for remote command but mishandles groups

\ } while (0) #endif On Mon, Jan 18, 2021 at 12:22 PM Ondrej Dubaj wrote: > When tar executes remote command for remote archive (by default > /usr/bin/ssh), it calls sys_reset_uid_gid in rmt_open__ in forked child right > before execl. So it seems like the intent

Re: tar attempts to drop privileges for remote command but mishandles groups

Agree, thanks for your solution! Will this solution be part of upstream code in near future ? Thanks. Ondrej On Mon, Feb 8, 2021 at 5:06 PM Sergey Poznyakoff wrote: > FWIW, I'd rather propose something like that: > > diff --git a/lib/system.h b/lib/system.h > index e7f531c..dffab86 100644 > -

Re: Disk read error interpreted as "File shrank" + not saying it is "Padding with zeros", but padds silently

Gentle ping On Mon, Jan 18, 2021 at 12:02 PM Ondrej Dubaj wrote: > One of the customer faced I/O errors while archiving a huge file 11 TB and > observed that after Tar had hit read I/O error due to xfs filesystem, it > still continue writing 0's to the file using strace. Howeve

Re: Disk read error interpreted as "File shrank" + not saying it is "Padding with zeros", but padds silently

Ping, any updates here? Thanks. On Mon, Feb 15, 2021 at 5:07 PM Ondrej Dubaj wrote: > Gentle ping > > On Mon, Jan 18, 2021 at 12:02 PM Ondrej Dubaj wrote: > >> One of the customer faced I/O errors while archiving a huge file 11 TB and >> observed that after Tar had h

Re: Disk read error interpreted as "File shrank" + not saying it is "Padding with zeros", but padds silently

Hello, any update here? Thanks. Ondrej On Mon, Mar 1, 2021 at 11:05 AM Ondrej Dubaj wrote: > Ping, any updates here? > > Thanks. > > On Mon, Feb 15, 2021 at 5:07 PM Ondrej Dubaj wrote: > >> Gentle ping >> >> On Mon, Jan 18, 2021 at 12:02 PM Ondrej Duba

Re: Disk read error interpreted as "File shrank" + not saying it is "Padding with zeros", but padds silently

gentle ping On Fri, Mar 19, 2021 at 12:22 PM Ondrej Dubaj wrote: > Hello, > any update here? > > Thanks. > Ondrej > > On Mon, Mar 1, 2021 at 11:05 AM Ondrej Dubaj wrote: > >> Ping, any updates here? >> >> Thanks. >> >> On Mon, Feb 15, 2

tar - covscan issues

Hello, proposing patch for some of the issues found by coverity scan in tar-1.34 Patch: diff --git a/gnu/malloc/scratch_buffer_dupfree.c b/gnu/malloc/scratch_buffer_dupfree.c index 775bff5..3b246f2 100644 --- a/gnu/malloc/scratch_buffer_dupfree.c +++ b/gnu/malloc/scratch_buffer_dupfree.c @@ -35,

cpio - covscan issues

Hello, proposing patch for some of the issues found by coverity scan in cpio-2.13 Patch: diff --git a/src/tar.c b/src/tar.c index 99ef8a2..a5873e7 100644 --- a/src/tar.c +++ b/src/tar.c @@ -146,6 +146,7 @@ write_out_tar_header (struct cpio_file_stat *file_hdr, int out_des) name_len = strlen (

Re: cpio - covscan issues

On Thu, Apr 8, 2021 at 12:02 PM Kamil Dudka wrote: > On Thursday, April 8, 2021 9:47:05 AM CEST Ondrej Dubaj wrote: > > Hello, > > > > proposing patch for some of the issues found by coverity scan in > cpio-2.13 > > > > Patch: > > > > diff --git a/

Re: Disk read error interpreted as "File shrank" + not saying it is "Padding with zeros", but padds silently

Hello, any updates here? Is this a real issue or a misunderstanding? Thank you. Regards, Ondrej On Tue, Mar 30, 2021 at 4:09 PM Ondrej Dubaj wrote: > gentle ping > > On Fri, Mar 19, 2021 at 12:22 PM Ondrej Dubaj wrote: > >> Hello, >> any update here? >> >