--- lib/pv_move.c 2003-07-11 19:10:26.000000000 -0300 +++ lib/pv_move.c.rcom 2003-07-11 19:09:35.000000000 -0300 @@ -916,7 +916,7 @@ ret =3D ioctl(group, PE_LOCKED_COPY, &pe_copy_req); if (ret < 0) ret =3D -errno; - if (ret =3D=3D -EINVAL) + if (ret =3D=3D -EINVAL || ret =3D=3D -ENOTTY) ret =3D -LVM_EPV_LOCKED_COPY_EINVAL; =09 close(group);

Thank you. Is that patch going to be added to the main source code base?

I think I found an error (segfault), see below.

==> I have a few more improvement suggestions.

* IMHO vgs/lvs should include a column called seg_size_pe, similar to seg_start and seg_start_pe
* IMHO vgs/lvs/pvs should have an command line parameter, or option modifier allowing to not display empty columns, i.e. the following command might skip displaying origin, snap%, move, log, copy%, convert, because they have no values. That would allow me to show copy_percent column only when copy is in progress. Would be nice.

lvs --all -o +vg_extent_size LV      VG     Attr   LSize   Origin Snap% Move Log Copy% Convert Ext   boot    volume -wi-ao   1.00g                                       4.00m copy    volume -wi-ao   2.00t                                       4.00m emulate volume -wi-ao 50.00g                                       4.00m home    volume -wi-ao 150.00g                                       4.00m p2p     volume -wi-a- 300.00g                                       4.00m root    volume -wi-ao   4.00g                                       4.00m swap    volume -wi-ao   8.00g                                       4.00m tmp     volume -wi-ao   2.00g                                       4.00m usr     volume -wi-ao 20.00g                                       4.00m var     volume -wi-ao 20.00g                                       4.00m

Possible error:

root@localhost# lvs --rows --alignDoes not align, should it?==> SEGFAULT INFO

root@localhost# pvs --all -o +vg_fmtSegmentation fault (core dumped)However the following seems to work..... root@localhost# pvs --all -o +vg_name PV         VG     Fmt Attr PSize PFree VG    /dev/dm-0              --      0     0        --- some entries removed for brevity ---- /dev/md127 volume lvm2 a-   5.46t 2.77t volume --- some entries removed for brevity ---- /dev/sde1              --      0     0       Additional information that might help.root@localhost# pvs --version LVM version:     2.02.66(2) (2010-05-20) Library version: 1.02.48 (2010-05-20) Driver version: 4.22.0root@localhost# cat /etc/lsb-release DISTRIB_ID=UbuntuDISTRIB_RELEASE=12.04DISTRIB_CODENAME=preciseDISTRIB_DESCRIPTION="Ubuntu 12.04.2 LTS"root@localhost# valgrind pvs --all -o +vg_fmt==14584== Memcheck, a memory error detector==14584== Copyright (C) 2002-2011, and GNU GPL'd, by Julian Seward et al.==14584== Using Valgrind-3.7.0 and LibVEX; rerun with -h for copyright info==14584== Command: pvs --all -o +vg_fmt==14584== ==14584== Invalid read of size 8==14584==    at 0x548ED77: dm_report_field_string (in /lib/libdevmapper.so.1.02.1)==14584==    by 0x549006F: dm_report_object (in /lib/libdevmapper.so.1.02.1)==14584==    by 0x46B138: report_object (in /sbin/lvm)==14584==    by 0x423618: ??? (in /sbin/lvm)==14584==    by 0x4254B2: process_each_pv (in /sbin/lvm)==14584==    by 0x4230D3: ??? (in /sbin/lvm)==14584==    by 0x41AE60: lvm_run_command (in /sbin/lvm)==14584==    by 0x41B307: lvm2_main (in /sbin/lvm)==14584==    by 0x58CE76C: (below main) (libc-start.c:226)==14584== Address 0x28 is not stack'd, malloc'd or (recently) free'd==14584== ==14584== ==14584== Process terminating with default action of signal 11 (SIGSEGV)==14584== Access not within mapped region at address 0x28==14584==    at 0x548ED77: dm_report_field_string (in /lib/libdevmapper.so.1.02.1)==14584==    by 0x549006F: dm_report_object (in /lib/libdevmapper.so.1.02.1)==14584==    by 0x46B138: report_object (in /sbin/lvm)==14584==    by 0x423618: ??? (in /sbin/lvm)==14584==    by 0x4254B2: process_each_pv (in /sbin/lvm)==14584==    by 0x4230D3: ??? (in /sbin/lvm)==14584==    by 0x41AE60: lvm_run_command (in /sbin/lvm)==14584==    by 0x41B307: lvm2_main (in /sbin/lvm)==14584==    by 0x58CE76C: (below main) (libc-start.c:226)==14584== If you believe this happened as a result of a stack==14584== overflow in your program's main thread (unlikely but==14584== possible), you can try to increase the size of the==14584== main thread stack using the --main-stacksize= flag.==14584== The main thread stack size used in this run was 8388608.==14584== root@localhost# ltrace pvs --all -o +vg_fmt 2>&1 | tail__ctype_toupper_loc()                            = 0x7f2f0e83f768__ctype_toupper_loc()                            = 0x7f2f0e83f768__ctype_toupper_loc()                            = 0x7f2f0e83f768__sprintf_chk(0x1e86e88, 1, -1, 0x48daa5, 0x48dab8) = 2dm_pool_strdup(0x1e88120, 0x1e86e88, 0x1e86e8a, 0, 0x48daa8) = 0x1e8e538dm_pool_alloc(0x1e88120, 8, 2, 0x7f2f0d892030, 0x48daa8) = 0x1e8e540dm_report_field_set_value(0x1e8e508, 0x1e8e538, 0x1e8e540, 0x7f2f0d892030, 0x48daa8) = 0x1e8e540dm_report_field_string(0x1e88080, 0x1e8e550, 40, 0x6b1fe0, 0x1e6cfe0 <unfinished ...>--- SIGSEGV (Segmentation fault) ---+++ killed by SIGSEGV +++
root@localhost#root@localhost# gdb --args pvs --all -o +vg_fmtGNU gdb (Ubuntu/Linaro 7.4-2012.04-0ubuntu2.1) 7.4-2012.04Copyright (C) 2012 Free Software Foundation, Inc.License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>This is free software: you are free to change and redistribute it.There is NO WARRANTY, to the extent permitted by law. Type "show copying"and "show warranty" for details.This GDB was configured as "x86_64-linux-gnu".For bug reporting instructions, please see:<http://bugs.launchpad.net/gdb-linaro/>...Reading symbols from /sbin/pvs...(no debugging symbols found)...done.(gdb) runStarting program: /sbin/pvs --all -o +vg_fmt[Thread debugging using libthread_db enabled]Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".File descriptor 3 (socket:[46992174]) leaked on pvs invocation. Parent PID 14623: gdbFile descriptor 4 (socket:[46992175]) leaked on pvs invocation. Parent PID 14623: gdbFile descriptor 5 (pipe:[46992176]) leaked on pvs invocation. Parent PID 14623: gdbFile descriptor 6 (pipe:[46992176]) leaked on pvs invocation. Parent PID 14623: gdbProgram received signal SIGSEGV, Segmentation fault.0x00007ffff757dd77 in dm_report_field_string () from /lib/libdevmapper.so.1.02.1(gdb) disassemble dm_report_field_stringDump of assembler code for function dm_report_field_string:   0x00007ffff757dd70 <+0>:    push   %rbx   0x00007ffff757dd71 <+1>:    mov    %rsi,%rbx   0x00007ffff757dd74 <+4>:    mov    (%rdi),%rdi=> 0x00007ffff757dd77 <+7>:    mov    (%rdx),%rsi   0x00007ffff757dd7a <+10>:    callq 0x7ffff7570a30 <dm_pool_strdup@plt>   0x00007ffff757dd7f <+15>:    test   %rax,%rax(gdb) info registers rax            0x6dd550    7198032rbx            0x6dd550    7198032rcx            0x6b1fe0    7020512rdx            0x28    40rsi            0x6dd550    7198032rdi            0x6d7120    7172384rbp            0x6d7080    0x6d7080

On 07/20/2013 02:20 AM, Alasdair G Kergon wrote:

It uses lists of ranges separated by colons. :x-y If x is missing, 0 is inserted; if y is missing, the last extent on the device is inserted. Man pages can always be improved with more examples! For +length, try the patch below. Alasdair --- a/tools/toollib.c +++ b/tools/toollib.c @@ -1001,7 +1001,7 @@ static int _parse_pes(struct dm_pool *mem, char *c, struct dm_list *pe_ranges, const char *pvname, uint32_t size) { char *endptr; - uint32_t start, end; + uint32_t start, end, len; /* Default to whole PV */ if (!c) { @@ -1041,7 +1041,16 @@ static int _parse_pes(struct dm_pool *mem, char *c, struct dm_list *pe_ranges, goto error; c = endptr; } + } else if (*c == '+') { /* Length? */ + c++; + if (isdigit(*c)) { + if (!xstrtouint32(c, &endptr, 10, &len)) + goto error; + c = endptr; + end = start + (len ? (len - 1) : 0); + } } + if (*c && *c != ':') goto error;

Juan Pablo Giménez