Date: Fri, 22 Oct 2010 10:01:51 -0700 From: Maksim Yevmenkin <maksim.yevmenkin@gmail.com> To: Iain Hibbert <plunky@rya-online.net> Cc: freebsd-bluetooth@freebsd.org Subject: Re: obexapp get failure Message-ID: <AANLkTi=iHT9WOVuNq_VFfWo6R3J1ynkhGBV0s6VkS_Uw@mail.gmail.com> In-Reply-To: <1287738768.915002.8520.nullmailer@galant.ukfsn.org> References: <1287509041.022618.4884.nullmailer@galant.ukfsn.org> <AANLkTimVZRNp8oh9%2BVQ6rPbXXFjdNqpgSQ2AqWjqATJY@mail.gmail.com> <1287561876.893861.6837.nullmailer@galant.ukfsn.org> <AANLkTi=V8pm51C4D58KRsiz0FfhOiU=NnUYGQJdeUwxA@mail.gmail.com> <1287732977.227959.8695.nullmailer@galant.ukfsn.org> <1287738768.915002.8520.nullmailer@galant.ukfsn.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, Oct 22, 2010 at 2:12 AM, Iain Hibbert <plunky@rya-online.net> wrote: > On Fri, 22 Oct 2010, Iain Hibbert wrote: > >> I have one bug-report that I haven't looked into properly, not sure if its >> obexapp or my phone (windows mobile 6) that is the problem.. if I connect >> to FTRN on phone, using folder browsing service and try GET on a larger >> file it fails and loses sync. I've done some tests and the largest file I >> can GET is 65528 bytes (can send larger files from either side no problem) >> >> for example with two files (that are text files full of 'a') >> >> % obexapp -a touch -C ftrn -f >> obex> ls >> Access Owner Group Size Modified Name >> .. >> n/a n/a 65528 n/a test.small >> n/a n/a 65529 n/a test.large >> Success, response: OK, Success (0x20) >> obex> get test.small >> 65528 bytes streamed in 4 seconds (16382 bytes/sec) >> Success, response: OK, Success (0x20) >> obex> get test.large >> Failure, response: Unknown response (0x53) >> obex> ls >> Access Owner Group Size Modified Name >> Could not parse XML: (null) >> Success, response: OK, Success (0x20) >> obex> ls >> Access Owner Group Size Modified Name >> .. >> n/a n/a 65528 n/a test.small >> n/a n/a 65529 n/a test.large >> Success, response: OK, Success (0x20) >> obex> get test.small >> 65528 bytes streamed in 3 seconds (21842 bytes/sec) >> Success, response: OK, Success (0x20) >> obex> dis >> Success, response: OK, Success (0x20) >> >> I guess its something to do with a uint16_t and a small amount of header >> but I've no time to look at that until early next week.. > > looking at the hcidump output for the test.large transaction, > > < ACL data: handle 13 flags 0x02 dlen 42 > L2CAP(d): cid 0x0089 len 38 [psm 3] > RFCOMM(d): UIH: cr 1 dlci 8 pf 1 ilen 33 fcs 0x46 credits 6 > OBEX: Get cmd(f): len 33 > Connection ID (0xcb) = 27 > Name (0x01) = Unicode length 22 > 0000: 00 74 00 65 00 73 00 74 00 2e 00 6c 00 61 00 72 .t.e.s.t...l.a.r > 0010: 00 67 00 65 00 00 .g.e.. > >> ACL data: handle 13 flags 0x02 dlen 11 > L2CAP(d): cid 0x0044 len 7 [psm 3] > RFCOMM(d): UIH: cr 0 dlci 8 pf 0 ilen 3 fcs 0x80 > OBEX: Get rsp(f): status 100 len 65535 > Status 100 = Continue > Body (0x48) = Sequence length 65529 > 0000: 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61 aaaaaaaaaaaaaaaa > 0010: 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61 aaaaaaaaaaaaaaaa > 0020: 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61 aaaaaaaaaaaaaaaa > [...] > ffc0: 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61 aaaaaaaaaaaaaaaa > ffd0: 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61 aaaaaaaaaaaaaaaa > ffe0: 61 61 61 61 61 61 61 61 61 0a 61 61 61 61 61 61 aaaaaaaaa.aaaaaa > fff0: 61 61 61 61 61 61 61 61 0a aaaaaaaa. > > < ACL data: handle 13 flags 0x02 dlen 12 > L2CAP(d): cid 0x0089 len 8 [psm 3] > RFCOMM(d): UIH: cr 1 dlci 8 pf 1 ilen 3 fcs 0x46 credits 6 > OBEX: Get cmd(f): len 3 (continue) > >> ACL data: handle 13 flags 0x02 dlen 11 > L2CAP(d): cid 0x0044 len 7 [psm 3] > RFCOMM(d): UIH: cr 0 dlci 8 pf 0 ilen 3 fcs 0x80 > OBEX: Get rsp(f): status 500 len 3 > > it seems that the wm6 phone tells us that more is to come, so we issue a > continuation get but the phone is only confused.. I don't know if this is > done correctly, or can be worked around? hmmm... interesting.... could you please try to lower obexapp mtu? i.e. use -m command line switch. try to set to something like 16k-32k and see if it makes any difference. there is an OBEXAPP_BUFFER_SIZE which is set to 65536 by default. but, i think, it might be wrong. it probably should be set to OBEX_MAXIMUM_MTU (which just happens to be set to 65535 in my version of openobex library). i did not make any changes in this area recently. i recall doing torture testing when i was transferring 4mb+ files between 2 unix machines and between unix and windows machine without any problems. thanks, max
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?AANLkTi=iHT9WOVuNq_VFfWo6R3J1ynkhGBV0s6VkS_Uw>