Package Details: dsview-git 1.3.2.r53.g2e9e2c8-18

Git Clone URL: https://aur.archlinux.org/dsview-git.git (read-only, click to copy)
Package Base: dsview-git
Description: GUI programe for supporting various instruments from DreamSourceLab, including logic analyzer, oscilloscope, etc.
Upstream URL: http://www.dreamsourcelab.com/
Keywords: electronic signal
Licenses: GPL-3.0-only
Conflicts: dsview, dsview-cli
Provides: dsview, dsview-cli
Submitter: anatolik
Maintainer: taotieren
Last Packager: taotieren
Votes: 4
Popularity: 0.000000
First Submitted: 2015-05-16 00:06 (UTC)
Last Updated: 2026-05-09 08:03 (UTC)

Latest Comments

« First ‹ Previous 1 2

FuzzyCheese commented on 2016-11-21 17:01 (UTC) (edited on 2016-11-21 17:29 (UTC) by FuzzyCheese)

Hangs on my Arch system and on a fresh Antergos install but works on Ubuntu fresh install. Symptom: Stays at 100% completion and UI is frozen and must be killed when used with DSLogic hardware. Crude fix: Comment out libusb event handling in receive_data function in libsigrok4DSL/hardware/DSL/dslogic.c around line 2365. --- a/DSView-master/libsigrok4DSL/hardware/DSL/dslogic.c +++ b/DSView-master/libsigrok4DSL/hardware/DSL/dslogic.c @@ -2380,11 +2380,11 @@ usb = sdi->conn; tv.tv_sec = tv.tv_usec = 0; - libusb_handle_events_timeout_completed(drvc->sr_ctx->libusb_ctx, &tv, &completed); +//fix libusb_handle_events_timeout_completed(drvc->sr_ctx->libusb_ctx, &tv, &completed); if (devc->status == DSL_FINISH) { - if (libusb_try_lock_events(drvc->sr_ctx->libusb_ctx) == 0) { - if (libusb_event_handling_ok(drvc->sr_ctx->libusb_ctx)) { +//fix if (libusb_try_lock_events(drvc->sr_ctx->libusb_ctx) == 0) { +//fix if (libusb_event_handling_ok(drvc->sr_ctx->libusb_ctx)) { /* Stop GPIF acquisition */ usb = ((struct sr_dev_inst *)devc->cb_data)->conn; if ((ret = command_stop_acquisition (usb->devhdl)) != SR_OK) @@ -2393,9 +2393,9 @@ sr_info("%s: Sent acquisition stop command!", __func__); remove_sources(devc); - } - libusb_unlock_events(drvc->sr_ctx->libusb_ctx); - } +//fix } +//fix libusb_unlock_events(drvc->sr_ctx->libusb_ctx); +//fix } } return TRUE; Details: The hang occurs because of libusb event handling with multiple threads.?.? The following line(113) in libsigrok4DSL/hardware/command.c does not return at the end of final transfer: ret = libusb_control_transfer(devhdl, LIBUSB_REQUEST_TYPE_VENDOR | LIBUSB_ENDPOINT_OUT, CMD_START, 0x0000, 0x0000, (unsigned char *)&cmd, sizeof(cmd), 3000); Used "export LIBUSB_DEBUG=4" to turn on libusb debug and "DSView -l 5" for application logging and it looks like the event is returned to another thread causing the current thread to wait on line 113. Log(trimmed): sr: finish_acquisition: send SR_DF_END packet sr: session: bus: Received SR_DF_END packet. [ 3.219458] [000042bf] libusb: debug [libusb_alloc_transfer] transfer 0x7f94fc000ed0 [ 3.219463] [000042bf] libusb: debug [libusb_submit_transfer] transfer 0x7f94fc000ed0 [ 3.219466] [000042bf] libusb: debug [add_to_flying_list] arm timerfd for timeout in 3000ms (first in line) [ 3.219491] [000042bf] libusb: debug [libusb_handle_events_timeout_completed] another thread is doing event handling ...repeats many times.... [ 6.298252] [000042be] libusb: debug [libusb_handle_events_timeout_completed] another thread is doing event handling [ 6.298330] [000042be] libusb: debug [libusb_cancel_transfer] transfer 0x7f94fc000ed0 [ 6.298338] [000042be] libusb: debug [discard_urbs] URB not found --> assuming ready to be reaped [ 6.298347] [000042be] libusb: debug [libusb_cancel_transfer] cancel transfer failed error -5 [ 6.298357] [000042be] libusb: warning [handle_timeout] async cancel failed -5 errno=22 [ 6.398431] [000042be] libusb: debug [libusb_handle_events_timeout_completed] another thread is doing event handling

anatolik commented on 2015-07-24 22:11 (UTC)

How do build the package? I have a fix for this problem in this package https://aur4.archlinux.org/cgit/aur.git/tree/0003-Fix-compile-error.patch?h=dsview-git so you should not see the problem.

Epileftric commented on 2015-07-24 22:08 (UTC)

There was an error when compiling, had to write the following patch http://pastebin.com/kw4wBTmm

« First ‹ Previous 1 2