Thread: Bug in new version of subsystem
Brought to you by:
aeb,
bencollins
From: Manfred W. <e95...@st...> - 2000-03-30 21:36:30
|
Hello! After updating to the new (cvs-)version of the subsystem (before I used ieee1394-2.2.14-20000206.gz) I observed the following very interesting phenomenon: If my computer was attached to a 1394-network during bootup (ie. when the driver was loaded) and I run testlibraw (from the current libraw1394 0.6 of course), it tells me that: <output of testlibraw> successfully got handle current generation number: 1 1 card(s) found nodes on bus: 0, card name: pcilynx using first card found: 0 nodes on bus, local ID is 3 doing transactions with custom tag handler using standard tag handler and synchronous calls testing FCP monitoring on local node got fcp command from node 3 of 8 bytes: 01 23 45 67 89 ab cd ef got fcp response from node 3 of 8 bytes: 01 23 45 67 89 ab cd ef polling for leftover messages </output of testlibraw> Obviously no node is found. But there must be at least one node (the local node). gscanbus crashes with a core dump (which probably is a bug of gscanbus which might not be able to handle 0 nodes correctly). Then I pulled the 1394 plug and afterwards it seemed to be ok, one node was detected (the same happens if the plug was already pulled during bootup): <output of testlibraw> successfully got handle current generation number: 2 1 card(s) found nodes on bus: 1, card name: pcilynx using first card found: 1 nodes on bus, local ID is 0 doing transactions with custom tag handler trying to send read request to node 0... completed with 0x00100000, value 0xfe060000 using standard tag handler and synchronous calls trying to read from node 0... completed with 0x00100000, value 0x77050000 testing FCP monitoring on local node got fcp command from node 0 of 8 bytes: 01 23 45 67 89 ab cd ef got fcp response from node 0 of 8 bytes: 01 23 45 67 89 ab cd ef polling for leftover messages </output of testlibraw> gscanbus does not core dump any more, but it says that it cannot read the topology map (this is probably because it is not yet adopted to the changed endianess policy). Afterwards I again plugged in the 1394-cable, and now the network is detected: <output of testlibraw> successfully got handle current generation number: 3 1 card(s) found nodes on bus: 9, card name: pcilynx using first card found: 9 nodes on bus, local ID is 3 doing transactions with custom tag handler trying to send read request to node 0... completed with 0x00020000, value 0x799b 96c6 trying to send read request to node 1... failed with error -1102 trying to send read request to node 2... completed with 0x00020000, value 0xa632 a2c6 trying to send read request to node 3... completed with 0x00100000, value 0x7205 0000 trying to send read request to node 4... completed with 0x00020000, value 0x3e23 bfc6 trying to send read request to node 5... failed with error -1102 trying to send read request to node 6... completed with 0x00020000, value 0xe4b3 76c6 trying to send read request to node 7... failed with error -1102 trying to send read request to node 8... failed with error -1102 using standard tag handler and synchronous calls trying to read from node 0... completed with 0x00020000, value 0x11b8acc6 trying to read from node 1... failed with error -1102 trying to read from node 2... completed with 0x00020000, value 0xf31ab8c6 trying to read from node 3... completed with 0x00100000, value 0xb00a0000 trying to read from node 4... completed with 0x00020000, value 0x82f4d4c6 trying to read from node 5... failed with error -1102 trying to read from node 6... completed with 0x00020000, value 0xd1828cc6 trying to read from node 7... failed with error -1102 trying to read from node 8... failed with error -1102 testing FCP monitoring on local node got fcp command from node 3 of 8 bytes: 01 23 45 67 89 ab cd ef got fcp response from node 3 of 8 bytes: 01 23 45 67 89 ab cd ef polling for leftover messages </output of testlibraw> It took me quite a lot of time to find out, that in order to work with the new driver, I have to pull the plug and then to re-plug-in. I hope, you will find the bug. Manfred Weihs PS: I can report another bug which I observed yesterday before updating, so it is a bug of the version from feb 6th, and I do not know if it still exists in the current version. If I plugged in the cabel, which a 1394-network hang on, and the X-server (Xfree86 3.6, kde 1.1.2) was running (and visible), then sometimes my pc hang up (completely, I was not even able to ping it). But it did not happen always, so I could not reproduce it. But it never happened when a text console was active, and booting while attached to the 1394-network also worked, it only hang up, when I plugged in. |