Re: [Qemu-devel] Regression: block: Add .bdrv_co_pwrite_zeroes()

[Peter Lieven]

Am 05.07.2016 um 03:53 schrieb Eric Blake:
> On 07/04/2016 07:49 AM, Peter Lieven wrote:
> > Hi,
> >
> > the above commit:
> >
> > commit d05aa8bb4a8b6aa9a915ec5074fb12ae632d2323
> > Author: Eric Blake <address@hidden>
> > Date:   Wed Jun 1 15:10:03 2016 -0600
> >
> >      block: Add .bdrv_co_pwrite_zeroes()
> >
> > introduces a regression (at least for me).
> >
> > The Limits from the iSCSI Block Limits VPD have no requirement of being
> > a power of two.
> > We use Dell Equallogic iSCSI SANs for instance. They have an internal
> > page size of 15MB. And
> > they advertise this page size as max_ws_len, opt_transfer_len and
> > opt_discard_alignment.
> A non-power-of-2 max_ws_len shouldn't be a problem, but opt_transfer_len
> and opt_discard_alignment not being a power of 2 impacts other code.
> 15MB is a rather odd page size.

I know, not my idea ;-) I think at least opt_discard_alignment of 15MB used to 
work
before.

> > I think we cannot assert that that these alignments are a power of 2.
> Perhaps that means we should just fix our code to round things down to
> the nearest power of 2 (8MB) for the opt_transfer_len and
> opt_discard_alignment values.  Can you post a stack-trace of the actual
> assertion you are hitting?

Sure:

qemu-system-x86_64: block/io.c:1165: bdrv_co_do_pwrite_zeroes: Assertion 
`is_power_of_2(alignment)' failed.

Program received signal SIGABRT, Aborted.
0x00007ffff5222c37 in __GI_raise (address@hidden) at 
../nptl/sysdeps/unix/sysv/linux/raise.c:56
56    ../nptl/sysdeps/unix/sysv/linux/raise.c: Datei oder Verzeichnis nicht 
gefunden.
(gdb) bt full
#0  0x00007ffff5222c37 in __GI_raise (address@hidden) at 
../nptl/sysdeps/unix/sysv/linux/raise.c:56
        resultvar = 0
        pid = 9610
        selftid = 9610
#1  0x00007ffff5226028 in __GI_abort () at abort.c:89
        save_stage = 2
        act = {__sigaction_handler = {sa_handler = 0x7fffffffe1fe, sa_sigaction 
= 0x7fffffffe1fe}, sa_mask = {__val = {140737307379972,
              93824998987040, 1165, 93823560581120, 140737306023251, 0, 
93825220359792, 47244640260, 93825009612448, 256, 0, 0, 0, 21474836480,
              140737354027008, 140737307395000}}, sa_flags = 1438406118, sa_restorer 
= 0x555555bc5ca0 <__PRETTY_FUNCTION__.34924>}
        sigs = {__val = {32, 0 <repeats 15 times>}}
#2  0x00007ffff521bbf6 in __assert_fail_base (fmt=0x7ffff536c3b8 "%s%s%s:%u: 
%s%sAssertion `%s' failed.\n%n",
    address@hidden "is_power_of_2(alignment)", address@hidden "block/io.c", 
address@hidden,
    address@hidden <__PRETTY_FUNCTION__.34924> "bdrv_co_do_pwrite_zeroes") at 
assert.c:92
        str = 0x5555586f5780 ""
        total = 4096
#3  0x00007ffff521bca2 in __GI___assert_fail (address@hidden 
"is_power_of_2(alignment)",
    address@hidden "block/io.c", address@hidden,
    address@hidden <__PRETTY_FUNCTION__.34924> "bdrv_co_do_pwrite_zeroes") at 
assert.c:101
No locals.
#4  0x0000555555a8a968 in bdrv_co_do_pwrite_zeroes (address@hidden, 
address@hidden, address@hidden,
    address@hidden) at block/io.c:1165
        drv = 0x5555560b3580 <bdrv_raw>
        qiov = {iov = 0x100000000, niov = 4096, nalloc = 0, size = 
140737148473344}
        iov = {iov_base = 0x0, iov_len = 0}
        ret = 0
        need_flush = false
        head = 0
        tail = 0
        max_write_zeroes = <optimized out>
        alignment = 15728640
        __PRETTY_FUNCTION__ = "bdrv_co_do_pwrite_zeroes"
#5  0x0000555555a8ae3b in bdrv_aligned_pwritev (address@hidden, address@hidden, 
address@hidden,
    address@hidden, qiov=0x555558909e58, flags=<optimized out>, address@hidden) 
at block/io.c:1290
---Type <return> to continue, or q <return> to quit---
        drv = 0x5555560b3580 <bdrv_raw>
        waited = <optimized out>
        ret = <optimized out>
        start_sector = 2656248
        end_sector = 2656256
        __PRETTY_FUNCTION__ = "bdrv_aligned_pwritev"
#6  0x0000555555a8b95f in bdrv_co_pwritev (bs=0x5555565b2df0, address@hidden, 
address@hidden,
    address@hidden, address@hidden) at block/io.c:1514
        req = {bs = 0x5555565b2df0, offset = 1359998976, bytes = 4096, type = 
BDRV_TRACKED_WRITE, serialising = false,
          overlap_offset = 1359998976, overlap_bytes = 4096, list = {le_next = 
0x55555a3c9cc0, le_prev = 0x5555565b5f28}, co = 0x555557ccc100,
          wait_queue = {entries = {tqh_first = 0x0, tqh_last = 
0x555562ee39b8}}, waiting_for = 0x0}
        align = 512
        head_buf = 0x0
        tail_buf = <optimized out>
        local_qiov = {iov = 0x0, niov = 1448841360, nalloc = 21845, size = 0}
        use_local_qiov = <optimized out>
        ret = <optimized out>
        __PRETTY_FUNCTION__ = "bdrv_co_pwritev"
#7  0x0000555555a7cae3 in blk_co_pwritev (blk=0x5555565b2c30, 
offset=1359998976, bytes=4096, qiov=0x555558909e58, flags=0)
    at block/block-backend.c:788
        ret = <optimized out>
#8  0x0000555555a7cc2e in blk_aio_write_entry (opaque=0x555557da5200) at 
block/block-backend.c:977
        acb = 0x555557da5200
        rwco = 0x555557da5228
#9  0x0000555555afafda in coroutine_trampoline (i0=<optimized out>, i1=<optimized 
out>) at util/coroutine-ucontext.c:78
        self = 0x555557ccc100
        co = 0x555557ccc100
#10 0x00007ffff5235800 in ?? () from /lib/x86_64-linux-gnu/libc.so.6
No symbol table info available.
#11 0x00007fffffffcf20 in ?? ()
No symbol table info available.
#12 0x0000000000000000 in ?? ()
No symbol table info available.

Peter