[sane-devel] sane-backends (eposn2) segfaults pretty much all the time
On Thu, Oct 3, 2013 at 4:32 AM, Johannes Bauer dfnsonfsduifb at gmx.de wrote: Hi list, I have two scanners, one that uses the epson2 backend and a nex ScanSnap ix500 which uses the fujitsu. For the newer ix500 I needed a fairly recent version, so I'm currently using git head of sane-backends. This report concerns 472daf78. First note: the epson2 backend doesn't compile, because the pedantic setup forces C90 and mixed declarations and code are C99. Change concerns backend/epson2-ops.c line 234 where I simply created a new scope before the declaration/initialization of int last until line 256. Then it compiles. However, it continuously blows up in my face with segfaults: $ scanimage --resolution 150 --mode Gray scanimage: output is not a file, exiting Segmentation fault $ scanimage --help Usage: scanimage [OPTION]... Start image acquisition on a scanner device and write image data to standard output. [...] Options specific to device `epson2:libusb:001:008': Scan Mode: --mode Lineart|Gray|Color [Lineart] [...] Type ``scanimage --help -d DEVICE'' to get list of all options for DEVICE. List of available devices: Segmentation fault This here starts scanning a complete document, actually creates a file, then outputs some broken debug output (control character is outputted) and then hangs (never exits): $ scanimage --resolution 150 --mode Gray x [Some weird CTRL char]libusb: 0.00 warning [libusb_close] internal signalling read failed, closing anyway libusb: 0.19 error [do_close] Device handle closed while transfer was still being processed, but the device is still connected as far as we know libusb: 0.26 error [do_close] A cancellation hasn't even been scheduled on the transfer for which the device is closing After I abort the hang with Ctrl-C, the scanner isn't recognized anymore and I need to plug it out and into USB again. Tracing the segfault for this command: $ /usr/local/bin/scanimage --resolution 150 --mode Gray yields: #0 0x74bd48c3 in libusb_submit_transfer () from /usr/lib64/libusb-1.0.so.0 #1 0x74bd600f in do_sync_bulk_transfer () from /usr/lib64/libusb-1.0.so.0 #2 0x74bd6384 in libusb_bulk_transfer () from /usr/lib64/libusb-1.0.so.0 #3 0x767fbf66 in usb_bulk_io.isra.1 () from /usr/lib64/libusb-0.1.so.4 #4 0x77bb93ae in sanei_usb_write_bulk (dn=0, buffer=0x7fffb760 \033Fz\365\377\177, size=0x7fffb708) at sanei_usb.c:2433 #5 0x742a3f3e in e2_send (s=0x627580, buf=0x7fffb760, buf_size=2, reply_len=4, status=0x7fffb75c) at epson2-io.c:99 #6 0x742a522d in esci_request_status (handle=0x627580, scanner_status=0x0) at epson2-commands.c:394 #7 0x7429ffc2 in close_scanner (s=0x627580) at epson2.c:365 #8 0x742a2214 in sane_epson2_close (handle=0x627580) at epson2.c:1545 #9 0x77bcb234 in sane_dll_close (handle=0x618710) at dll.c:1223 #10 0x00405d6d in scanimage_exit () at scanimage.c:1654 #11 0x75445721 in __run_exit_handlers (status=1, listp=0x757ad5a8 __exit_funcs, run_list_atexit=true) at exit.c:78 #12 0x754457a5 in __GI_exit (status=optimized out) at exit.c:100 #13 0x00402ad3 in main (argc=optimized out, argv=0x7fffd9f8) at scanimage.c:2040 The backtrace with the appropriate code: (gdb) up #1 0x74bd600f in do_sync_bulk_transfer () from /usr/lib64/libusb-1.0.so.0 (gdb) up #2 0x74bd6384 in libusb_bulk_transfer () from /usr/lib64/libusb-1.0.so.0 (gdb) up #3 0x767fbf66 in usb_bulk_io.isra.1 () from /usr/lib64/libusb-0.1.so.4 (gdb) up #4 0x77bb93ae in sanei_usb_write_bulk (dn=0, buffer=0x7fffb820 \033Fz\365\377\177, size=0x7fffb7c8) at sanei_usb.c:2433 2433 write_size = usb_bulk_write (devices[dn].libusb_handle, (gdb) up #5 0x742a3f3e in e2_send (s=0x627580, buf=0x7fffb820, buf_size=2, reply_len=4, status=0x7fffb81c) at epson2-io.c:99 99 *status = sanei_usb_write_bulk(s-fd, buf, n); (gdb) up #6 0x742a522d in esci_request_status (handle=0x627580, scanner_status=0x0) at epson2-commands.c:394 394 e2_send(s, params, 2, 4, status); (gdb) up #7 0x7429ffc2 in close_scanner (s=0x627580) at epson2.c:365 365 esci_request_status(s, NULL); (gdb) up #8 0x742a2214 in sane_epson2_close (handle=0x627580) at epson2.c:1545 1545close_scanner(s); (gdb) up #9 0x77bcb234 in sane_dll_close (handle=0x618710) at dll.c:1223 1223 (*(op_close_t)s-be-op[OP_CLOSE]) (s-handle); (gdb) up #10 0x00405d6d in scanimage_exit () at scanimage.c:1654 1654 sane_close (device); (gdb) up #11 0x75445721 in __run_exit_handlers (status=1, listp=0x757ad5a8 __exit_funcs, run_list_atexit=true) at exit.c:78 78cxafct (f-func.cxa.arg, status); (gdb)
[sane-devel] sane-backends (eposn2) segfaults pretty much all the time
On Thu, Oct 3, 2013 at 1:34 PM, Johannes Bauer dfnsonfsduifb at gmx.de wrote: On 03.10.2013 13:12, m. allan noah wrote: I would recommend that you use git bisect (or a manual equivalent) to find the commit that breaks epson2. Then we can investigate a fix. This is really messed up: 206e6ed094c080be903b154ea3d016cc226feb89 is the first bad commit commit 206e6ed094c080be903b154ea3d016cc226feb89 Author: m. allan noah kitno-guest at alioth.debian.org Date: Sat Jun 30 21:28:14 2012 -0400 Adding new kodakaio backend The commit 206e6e is REALLY the first bad one (the one where it starts segfaulting). Curiously there are no changes in epson2 in there. I confirmed that the predecessor, df6a8ef89dbd061fb383722ddd6272e07fc468b3, is working (somewhat). At least it is NOT segfaulting, but displaying in the end: [...] Type ``scanimage --help -d DEVICE'' to get list of all options for DEVICE. List of available devices: libusb: 0.00 error [do_close] Device handle closed while transfer was still being processed, but the device is still connected as far as we know libusb: 0.43 error [do_close] A cancellation hasn't even been scheduled on the transfer for which the device is closing (then it hangs indefinitely, as described before) No idea what the hell is going on here. Best regards, Johannes That info isn't to help on its own. If its really calling scanimage.c:2040 thats a simple function thats unlikely to fail. I suspect one of scanner drivers is corrupting memory while auto-detecting scanners and epson2 is the victim... and if removing the kodakaio driver improves things then maybe its that driver. Can you try editing your /etc/sane.d/dll.conf and commenting out all backends except epson2 and see if things work different? If the bisect is pointing towards kodakaio driver then you can also try only commenting out that driver to try to narrow down to it. Chris
[sane-devel] Fwd: Schedule for release of sane-backends 1.0.24
I've completed a successful test run (make distcheck) on Fedora 20, OS X 10.5, and Windows/mingw32 (only epson2 backend enabled on mingw32). Chris On Mon, Sep 23, 2013 at 4:21 PM, m. allan noah kitno455 at gmail.com wrote: I'll probably have to slip the release back one day anyway. allan On Mon, Sep 23, 2013 at 4:40 PM, Chris Bagwell chris at cnpbagwell.com wrote: Kinda snuck up on me. I probably won't be able to do another test compile on OSX, Windows, and Fedora until towards end of week. Chances are high that they are still compiling since last month's test run though. Chris On Mon, Sep 23, 2013 at 3:25 PM, m. allan noah kitno455 at gmail.com wrote: Yes. But, I'm not fanatical about it. If you have something simple that you are sure won't break anything else, go ahead and commit it. allan On Mon, Sep 23, 2013 at 4:09 PM, Rolf Bensch rolf at bensch-online.de wrote: Are we in code freeze now? Rolf Original-Nachricht Betreff: [sane-devel] Schedule for release of sane-backends 1.0.24 Datum:Sat, 24 Aug 2013 13:16:08 -0400 Von: m. allan noah kitno455 at gmail.com An: sane-devel sane-devel at lists.alioth.debian.org It has been 1 year since our last release, so here we go again: Timetable: Sept 15, 2013: Feature freeze (only bugs, translation and doc updates) Sept 22, 2013: Code freeze (only horrible bugs, translation and doc updates) Sept 29, 2013: Release Note, it is highly likely that there are patches from end users which we failed to apply. If you know of such a case, please speak up, so we can start a dialog about including them in the release. During the period leading up to the release, it would be helpful if we could get some builds on various platforms, and tests done with various scanners. The sooner we discover problems, the better. allan -- The truth is an offense, but not a sin -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20130926/a3aaeb64/attachment.html
[sane-devel] Fwd: Schedule for release of sane-backends 1.0.24
Kinda snuck up on me. I probably won't be able to do another test compile on OSX, Windows, and Fedora until towards end of week. Chances are high that they are still compiling since last month's test run though. Chris On Mon, Sep 23, 2013 at 3:25 PM, m. allan noah kitno455 at gmail.com wrote: Yes. But, I'm not fanatical about it. If you have something simple that you are sure won't break anything else, go ahead and commit it. allan On Mon, Sep 23, 2013 at 4:09 PM, Rolf Bensch rolf at bensch-online.de wrote: Are we in code freeze now? Rolf Original-Nachricht Betreff: [sane-devel] Schedule for release of sane-backends 1.0.24 Datum:Sat, 24 Aug 2013 13:16:08 -0400 Von: m. allan noah kitno455 at gmail.com An: sane-devel sane-devel at lists.alioth.debian.org It has been 1 year since our last release, so here we go again: Timetable: Sept 15, 2013: Feature freeze (only bugs, translation and doc updates) Sept 22, 2013: Code freeze (only horrible bugs, translation and doc updates) Sept 29, 2013: Release Note, it is highly likely that there are patches from end users which we failed to apply. If you know of such a case, please speak up, so we can start a dialog about including them in the release. During the period leading up to the release, it would be helpful if we could get some builds on various platforms, and tests done with various scanners. The sooner we discover problems, the better. allan -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20130923/8f2d086a/attachment.html
[sane-devel] sanei_usb_open() and permission denied
Ok, I pushed a patch that lets USB test pass on all my boxes. I'd appreciate if someone could test this (run make check) on a box with a USB scanner plugged in. All my scanners are network scanners so I can't test that case. Chris On Sun, Aug 25, 2013 at 12:58 PM, m. allan noah kitno455 at gmail.com wrote: Your suggested changes sound fine to me. allan On Sun, Aug 25, 2013 at 1:09 PM, Chris Bagwell chris at cnpbagwell.com wrote: Hi all, This is probably directly more towards St?phane based on recent USB work but of interest to all. The sanie_usb_test.c test is failing on one of my Linux boxes in the test_open_all() test... I'd like to get some input on best way to fix this failure or if we should fix it at all. The reason it is failing for me is because this box has a USB wifi module that has a report with Vendor Specific Class/Subclass/Protocol and so sane adds it to the list as possible scanner. The open itself fails when ran by normal user because its accessable only by root (EACCES). The open fails when ran as root because its already opened by networking stuff (EBUSY). Because these two cases are kinda soft failures, I was thinking about changing the test to simply skip over these opens instead of failing the whole test case. What do you think about this change? If you agree with that change, the second issue I'll have is that sanei_usb_open() seems to treat EPERM as permission denied (SANE_STATUS_ACCESS_DENIED) but not EACCES. Do you see any issues with me changing it so both EPERM and EACCES map to SANE_STATUS_ACCESS_DENIED? If I don't make that change then the EACCES error code is mapped to generic SANE_STATUS_INVAL and I can't tell difference between real failures and permission failures. Chris -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org -- The truth is an offense, but not a sin -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20130826/5e9333ff/attachment.html
[sane-devel] epson2 backend URL fixes (was: Re: Epson GT-7300U)
Applied this patch and your other model patch to git. Someone else will have to do the HTML part. Chris On Sun, Aug 18, 2013 at 10:41 PM, Olaf Meeuwissen olaf.meeuwissen at avasys.jp wrote: Just pinging the list about the status of the patch I sent mid June. Anyone willing to commit? Olaf Meeuwissen writes: Julian H. Stacey writes: To whoever maintains page: http://www.sane-project.org/sane-backends.html At line found by GT-7300U The URL behind is http://avasys.jp/product/linux_driver/index.php which needs changing as - Its Japanese (with an English button) to http://avasys.jp/eng/ the their: 2013/01/21Notice of termination of the Linux Driver Download Service So now please point to http://download.ebz.epson.net/dsc/search/01/search/searchModule Looks like Julian is talking about the information from the epson2 backend. Please find a patch attached that fixes all the URLs. If someone with commit privileges could push this (and update the HTML pages) I'd appreciate it. The patch can be applied with `git am`. Hope this helps, -- Olaf Meeuwissen, LPIC-2 FLOSS Engineer -- AVASYS CORPORATION FSF Associate Member #1962 Help support software freedom http://www.fsf.org/jf?referrer=1962 -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20130825/a7067494/attachment.html
[sane-devel] sanei_usb_open() and permission denied
Hi all, This is probably directly more towards St?phane based on recent USB work but of interest to all. The sanie_usb_test.c test is failing on one of my Linux boxes in the test_open_all() test... I'd like to get some input on best way to fix this failure or if we should fix it at all. The reason it is failing for me is because this box has a USB wifi module that has a report with Vendor Specific Class/Subclass/Protocol and so sane adds it to the list as possible scanner. The open itself fails when ran by normal user because its accessable only by root (EACCES). The open fails when ran as root because its already opened by networking stuff (EBUSY). Because these two cases are kinda soft failures, I was thinking about changing the test to simply skip over these opens instead of failing the whole test case. What do you think about this change? If you agree with that change, the second issue I'll have is that sanei_usb_open() seems to treat EPERM as permission denied (SANE_STATUS_ACCESS_DENIED) but not EACCES. Do you see any issues with me changing it so both EPERM and EACCES map to SANE_STATUS_ACCESS_DENIED? If I don't make that change then the EACCES error code is mapped to generic SANE_STATUS_INVAL and I can't tell difference between real failures and permission failures. Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20130825/0d958d4d/attachment.html
[sane-devel] Compiling question (autoreconf)
On Fri, Aug 23, 2013 at 10:46 AM, ?l ??gn?r sane-ml at corr.eu.org wrote: I compiled sane from git as mentioned below. Can I ignore the messages with autoreconf? $ cd /install/linuxins/sane-git/src/ $ git clone git://git.debian.org/sane/sane-backends.git $ cd /install/linuxins/sane-git/src/sane-backends/ Because all the autofoo files are committed to git, after the clone you should be able to skip all the next steps and jump straight to the ./configure step. You only need to do these steps with warnings if you mess with configure.in or Makefile.am's. $ autoconf aclocal.m4:17: warning: this file was generated for autoconf 2.69. You have another version of autoconf. It may work, but is not guaranteed to. If you have problems, you may need to regenerate the build system entirely. To do so, use the procedure documented by the package, typically 'autoreconf'. autoreconf is same as running both autoconf and automake so you shouldn't be doing all three commands. $ autoreconf libtoolize: Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am. backend/Makefile.am:52: `%'-style pattern rules are a GNU make extension po/Makefile.am:17: shell if test -r $(srcdir: non-POSIX variable name po/Makefile.am:17: (probably a GNU make extension) po/Makefile.am:18: shell if test -r $(srcdir: non-POSIX variable name po/Makefile.am:18: (probably a GNU make extension) po/Makefile.am:20: shell LINGUAS=$(PO_LINGUAS: non-POSIX variable name po/Makefile.am:20: (probably a GNU make extension) po/Makefile.am:21: shell LINGUAS=$(PO_LINGUAS: non-POSIX variable name po/Makefile.am:21: (probably a GNU make extension) tools/Makefile.am:62: wildcard ${top_srcdir}/doc/descriptions/*.desc: non-POSIX variable name tools/Makefile.am:62: (probably a GNU make extension) tools/Makefile.am:62: wildcard ${top_srcdir}/doc/descriptions-external/*.desc: non-POSIX variable name tools/Makefile.am:62: (probably a GNU make extension) tools/Makefile.am:66: wildcard ${top_srcdir}/doc/descriptions/*.desc: non-POSIX variable name tools/Makefile.am:66: (probably a GNU make extension) tools/Makefile.am:66: wildcard ${top_srcdir}/doc/descriptions-external/*.desc: non-POSIX variable name tools/Makefile.am:66: (probably a GNU make extension) tools/Makefile.am:70: wildcard ${top_srcdir}/doc/descriptions/*.desc: non-POSIX variable name tools/Makefile.am:70: (probably a GNU make extension) tools/Makefile.am:70: wildcard ${top_srcdir}/doc/descriptions-external/*.desc: non-POSIX variable name tools/Makefile.am:70: (probably a GNU make extension) tools/Makefile.am:74: wildcard ${top_srcdir}/doc/descriptions/*.desc: non-POSIX variable name tools/Makefile.am:74: (probably a GNU make extension) tools/Makefile.am:74: wildcard ${top_srcdir}/doc/descriptions-external/*.desc: non-POSIX variable name tools/Makefile.am:74: (probably a GNU make extension) Yes, all these warnings are currently expected upon a successful run of autoreconf command. So it looked like this works for you if you do need to modify autofoo related files. Chris $ autoconf $ automake # Test # $BACKENDS=pixma ./configure # - The following backends will be built: # pixma $ ./configure $ make su - cd /install/linuxins/sane-git/src/sane-backend make install -- ?l -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20130823/6b538548/attachment.html
[sane-devel] Time to set up SANE's release plan ?
On Sun, Aug 11, 2013 at 3:31 PM, Stef stef.dev at free.fr wrote: Hello, it's about one year that SANE 1.0.23 has been released. I think we should set a release plan to get 1.0.24 ready this year. Regards, Stef For what its worth, I've finally submitted the one patch I wanted to get in before another release (a pthreads patch to work with mingw64) so I'm ready. Also, I'd like to get make distcheck working before next release if possible. Some of the new testsuite work is failing when distcheck runs the check target. I may be able to help debug that deeper at later date. For now, here is what I found. sanei_config_test is failing because its looking for a config file relative to current directory but that doesn't work when you build and run from a directory different from source. We somehow need to pass $srcdir during compiling and have it prepend that to config filename instead of looking in ./data. Also, sanei_usb_test is failing for me when ran from _build as well. I'm not sure the reason it works differently but I see that test_store_device() is calling store_device() and there is a for() loop that checks each value for devices[i]. A call to strcpy(devices[i].devname, device.devname) is crashing because devices[i].devname is a NULL pointer (whole structure except for devices[i].method are null vaules as matter of fact). Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20130815/fdba617b/attachment.html
[sane-devel] sane under windows XP / 7
On Wed, Sep 5, 2012 at 5:12 AM, Matteo Gasparin email.diversa at gmail.com wrote: I'm using the 1.0.22 tarball downloaded from https://alioth.debian.org/frs/?group_id=30186 You'll want to download the just released 1.0.23 or grab from git. It has several mingw32 bugfixes. For the environment variables, do I just need to set them and they will be picked by the dll? yes. See 'man sane-usb' and 'man sane-avision' for examples. Is the use of libdwi the only way to see the usb devices from Windows? Seems kind of a complicated and hackish way to get it work... What is libdwi? libusb-win32 is needed and I hear windows support has been integrated into libusb-1.x library directory but requires minor header file fixes in sane to use them. Chris
[sane-devel] sane under windows XP / 7
On Tue, Sep 4, 2012 at 10:33 AM, Matteo Gasparin email.diversa at gmail.com wrote: I'm trying to write a multi platform scanning utility using SANE but, while everything with linux is fine, opening the device and all, I find issues with windows, both xp 32 and 7 64 bit. I downloaded the source and compiled it using cygwin and the libusb-win32 library and everything went fine; some of the backends didn't build: *** canon_pp backend requires libieee1284 library - disabling *** gphoto2 backend requires gphoto2 and JPEG libraries - disabling *** hpsj5s backend requires libieee1284 library - disabling *** mustek_pp backend requires libieee1284 and paraport-directio libraries - disabling *** mustek_usb2 backend requires pthread library - disabling *** pint backend requires sys/scanio.h - disabling *** qcam backend requires ioperm and portaccess functions - disabling *** v4l backend requires v4l libraries - disabling *** pnm backend not manually selected - disabling which is ok for me, since the scanner I'm trying to reach is a HP Scanjet 5300C, which should use the avisio backend. I thought it could be a USB driver issue so I used libwdi's zadig program to switch the normal xp driver with the libusb-win32 one. After that sane-find-scanner could correctly recognize the scanner but no luck using scanimage. Did I do something wrong? Did I miss something? Hopefully, others with mingw32 experience will jump in as well. But to start, what version of sane-backends are you testing with? I occasionally test cross compiling sane-backends on Fedora using ming32 cross compiler and then I test the executable using wine with a networked Epson scanner. I compile with configure BACKENDS=epson2 so I don't get much visibility into other backends. I have no experience testing USB support under mingw32 though. I believe avision is the only other backend that has much mingw32 testing by other users. I thought there has been successful USB access using libusb-win32 though. Your probably going to have to start debugging to get much further. You can try the SANE_DEBUG_SANEI_USB and SANEI_DEBUG_AVISION environment variables and see if you can spot were things are going wrong between Linux and Windows. On top of that I'm using Java and JNA to interface with SANE; again everything using linux is fine, while with windows it correctly loads the library and then hangs at the first call of sane_init(null, null) . Someone got an idea about that? Hmm, not sure. I recall having some odd issues with finding config files on windows. I think you may have to change into scanimage install directory before running your app or set the SANE_CONFIG_DIR environment variable. That type of failures shouldn't cause hang ups but you never know. Chris
[sane-devel] sane-backends from git 2012-08-16
Thanks for the report and sorry about that. Second time I messed up with this change. I just submitted a fix for it. Its really odd how the different compiler versions and --disable-shared are all catching different symbol issues. My Fedora 17 wasn't finding this issue. This time, I ran configure BACKEND=kvsXXX for each of the 3 kvs backends I modified. Hopefully, that will have cought any remaining missing symbols. FYI: The Epjitsu and Snapscan backends both have a duplicate scan() symbol and what my compiler was using to not fail since I was linking. Its probably best to add prefixes to those backends at some point as well to prevent any issues in the off chance both backends are used on 1 system. Chris On Fri, Aug 17, 2012 at 1:39 PM, m. allan noah kitno455 at gmail.com wrote: There was a recent patch to kvs backends to allow them to build statically, by renaming some functions. It appears that one was missed. Try changing the call to kvs40xx_scan instead, and see if that builds. allan On Fri, Aug 17, 2012 at 2:24 PM, Sergei LITVINENKO sergei.litvinenko at gmail.com wrote: Hello everybody, I try to build rpm for F16 (spec from src.rpm package). In spec sane-backend is configured as: %configure \ --with-gphoto2=%{_prefix} \ --with-docdir=%{_docdir}/%{name}-%{version} \ --disable-locking --disable-rpath \ --enable-pthread Try to use like the same ./configure --with-gphoto2 --disable-locking --disable-rpath --enable-pthread and have: /bin/sh ../libtool --silent --tag=CC --mode=link gcc -g -O2 -W -Wall -Wcast-align -Wcast-qual -Wmissing-declarations -Wmissing-prototypes -Wpointer-arith -Wreturn-type -Wstrict-prototypes -pedantic -ansi -Wl,-z,defs -o libkvs40xx.la libkvs40xx_la-kvs40xx.lo libkvs40xx_la-kvs40xx_cmd.lo libkvs40xx_la-kvs40xx_opt.lo /bin/sh ../libtool --silent --tag=CC --mode=link gcc -g -O2 -W -Wall -Wcast-align -Wcast-qual -Wmissing-declarations -Wmissing-prototypes -Wpointer-arith -Wreturn-type -Wstrict-prototypes -pedantic -ansi -Wl,-z,defs -rpath '/usr/local/lib/sane' -version-number 1:0:23 -module -o libsane-kvs40xx.la libsane_kvs40xx_la-kvs40xx-s.lo ../lib/liblib.la libkvs40xx.la ../sanei/sanei_init_debug.lo ../sanei/sanei_constrain_value.lo ../sanei/sanei_config.lo sane_strstatus.lo ../sanei/sanei_usb.lo ../sanei/sanei_scsi.lo -lusb -lpthread ./.libs/libkvs40xx.a(libkvs40xx_la-kvs40xx.o): In function `sane_kvs40xx_start': /home/sergeil/desk/src/BUILD/sane-backends-20120816/backend/kvs40xx.c:595: undefined reference to `scan' collect2: ld returned 1 exit status make[2]: *** [libsane-kvs40xx.la] ?? 1 make[2]: ? ?? `/home/sergeil/desk/src/BUILD/sane-backends-20120816/backend' make[1]: *** [all] ?? 2 make[1]: ? ?? `/home/sergeil/desk/src/BUILD/sane-backends-20120816/backend' make: *** [all-recursive] ?? 1 Do you know where function scan() is defined? /home/sergeil/desk/src/BUILD/sane-backends-20120816/backend/kvs40xx.c ... if (duplex) { st = kvs40xx_set_window (s, SIDE_BACK); if (st) return st; } st = scan (s); if (st) return st; ... PS: In case of simple configuration `./configure`, make can finish OK. -- Best regards, Sergei LITVINENKO There is a great difference between be involved and be committed. Usually we are committed... -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org -- The truth is an offense, but not a sin -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org
[sane-devel] error building using --enable-static
Could you submit the changes then? With that combo below, it should make it so only frontends/Makefile.am and frontsends/Makefile.in need to be submitted that have minor changes. With my installed tools, its closer to 10 or more files with large changes. Chris On Wed, Aug 8, 2012 at 2:45 PM, Stef stef.dev at free.fr wrote: Hello, on my test system (Slackware 13.37): - i cloned current git tree - did a static build that failed linking - applied the library order patch - done 'autoreconf' - rebuilt successfully and got a ~9.5 MB scanimage and saned The 'auto' packages used were: autoconf 2.68 automake 1.11.1 So I believe it would be fine to apply it to the git tree. Regards, Stef
[sane-devel] error building using --enable-static
Thank you much, Allan. Although its compiling for me on OS X now, Stef identified an additional patch needed for Linux. Can you commit it as well? diff --git a/frontend/Makefile.am b/frontend/Makefile.am index 4ceae3b..74e7425 100644 --- a/frontend/Makefile.am +++ b/frontend/Makefile.am @@ -17,7 +17,8 @@ endif AM_CPPFLAGS = -I. -I$(srcdir) -I$(top_builddir)/include -I$(top_srcdir)/include scanimage_SOURCES = scanimage.c stiff.c stiff.h -scanimage_LDADD = ../backend/libsane.la ../lib/liblib.la ../lib/libfelib.la +scanimage_LDADD = ../backend/libsane.la ../sanei/libsanei.la ../lib/liblib.la \ + ../lib/libfelib.la saned_SOURCES = saned.c saned_LDADD = ../backend/libsane.la ../sanei/libsanei.la ../lib/liblib.la \ Chris On Thu, Aug 9, 2012 at 8:04 PM, m. allan noah kitno455 at gmail.com wrote: I have committed these changes. allan On Thu, Aug 9, 2012 at 12:52 PM, Chris Bagwell chris at cnpbagwell.com wrote: Could you submit the changes then? With that combo below, it should make it so only frontends/Makefile.am and frontsends/Makefile.in need to be submitted that have minor changes. With my installed tools, its closer to 10 or more files with large changes. Chris On Wed, Aug 8, 2012 at 2:45 PM, Stef stef.dev at free.fr wrote: Hello, on my test system (Slackware 13.37): - i cloned current git tree - did a static build that failed linking - applied the library order patch - done 'autoreconf' - rebuilt successfully and got a ~9.5 MB scanimage and saned The 'auto' packages used were: autoconf 2.68 automake 1.11.1 So I believe it would be fine to apply it to the git tree. Regards, Stef -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org -- The truth is an offense, but not a sin
[sane-devel] error building using --enable-static
On Wed, Aug 8, 2012 at 12:49 AM, Stef stef.dev at free.fr wrote: Hello, thanks for the patches. They work great here. For the linking problem, what should be the version number for old autoconf tools? The Makefile.in in git have a automake 1.11.1 tag, is it old enough? I may have a suitable test platform. Yes, 1.11.1 is ideal. I believe the current configure/Makefile.in's were generated based on versions in Fedora 16 if that helps as well. I have automake 1.11.3 from Fedora 17 and the changes are quite large for a minor release. To get scanimage compiling, I also had to specify libsanei in scanimage_LDADD: Oh yes, I remember that now. The info got lost because I was hand editing Makefile originally. Thanks, Chris
[sane-devel] error building using --enable-static
I fixed the kvs backend symbol issues and committed them. Also, I believe I have a fix for jpeg symbols. I think its just a linking ordering issue. At least I got it working by hand editing the frontend/Makefile and moving the libraries with missing symbols to later in line of saned_LDADD and scanimage_LDADD. I do not have the older versions of autofoo packages installed and I'll reluctant to submit large configure and Makefile.in file updates at this stage. If someone has older autofoo installed and can check in just frontend/Makefile.am and frontend/Makefile.in change, here is suggested patch (I notice that liblib.la is in two different orders. May need to move one of those to later as well): --- a/frontend/Makefile.am +++ b/frontend/Makefile.am @@ -17,11 +17,11 @@ endif AM_CPPFLAGS = -I. -I$(srcdir) -I$(top_builddir)/include -I$(top_srcdir)/include scanimage_SOURCES = scanimage.c stiff.c stiff.h -scanimage_LDADD = ../lib/liblib.la ../lib/libfelib.la ../backend/libsane.la +scanimage_LDADD = ../backend/libsane.la ../lib/liblib.la ../lib/libfelib.la saned_SOURCES = saned.c -saned_LDADD = ../sanei/libsanei.la ../lib/liblib.la ../lib/libfelib.la \ - ../backend/libsane.la @SYSLOG_LIBS@ +saned_LDADD = ../backend/libsane.la ../sanei/libsanei.la ../lib/liblib.la \ + ../lib/libfelib.la @SYSLOG_LIBS@ test_SOURCES = test.c test_LDADD = ../lib/liblib.la ../lib/libfelib.la ../backend/libsane.la Chris On Sat, Aug 4, 2012 at 2:13 AM, stef stef.dev at free.fr wrote: Hello, while debugging saned/genesys backend I noted that a static build : ./configure 'CFLAGS=-ggdb' '--prefix=/usr' '--sysconfdir=/etc' '--enable-pnm-backend' '--enable-parport-directio' '--enable-locking' --enable-static --disable-shared fails with: /usr/lib/gcc/i486-slackware-linux/4.7.1/../../../../i486-slackware-linux/bin/ld: Warning: alignment 16 of symbol `scan' in ../backend/.libs/libsane.a(libkvs20xx_la-kvs20xx_cmd.o) is smaller than 32 in ../backend/.libs/libsane.a(libhp3900_la-hp3900.o) /usr/lib/gcc/i486-slackware-linux/4.7.1/../../../../i486-slackware-linux/bin/ld: Warning: size of symbol `scan' changed from 72 in ../backend/.libs/libsane.a(libhp3900_la-hp3900.o) to 86 in ../backend/.libs/libsane.a(libkvs20xx_la-kvs20xx_cmd.o) /usr/lib/gcc/i486-slackware-linux/4.7.1/../../../../i486-slackware-linux/bin/ld: Warning: type of symbol `scan' changed from 1 to 2 in ../backend/.libs/libsane.a(libkvs20xx_la-kvs20xx_cmd.o) ../backend/.libs/libsane.a(libgphoto2_i_la-gphoto2.o): In function `converter_init': /stefdev/git/sane-backends/backend/gphoto2.c:1982: undefined reference to `sanei_jpeg_jinit_write_ppm' ../backend/.libs/libsane.a(libdc210_la-dc210.o): In function `sane_dc210_start': /stefdev/git/sane-backends/backend/dc210.c:1264: undefined reference to `sanei_jpeg_jinit_write_ppm' ../backend/.libs/libsane.a(libdc240_la-dc240.o): In function `sane_dc240_start': /stefdev/git/sane-backends/backend/dc240.c:1477: undefined reference to `sanei_jpeg_jinit_write_ppm' collect2: error: ld returned 1 exit status make[1]: *** [scanimage] Erreur 1 make: *** [all-recursive] Erreur 1 All the ksv20xx_cmd.c functions should be renamed like explained in backend-writing.txt , so to avoid clashes like the one on the 'scan' function with the hp3900 (which maybe is lacking a 'static' somewhere). 'make libchek' lists all the symbols that may clash, and there is quite a number of them. And it seems some library is missing at link time for scanimage. I currently haven't the time to dig into it. So if anyone has time Regards, Stef -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org
[sane-devel] Schedule for release of sane-backends 1.0.23
On Sun, Jul 29, 2012 at 7:26 PM, m. allan noah kitno455 at gmail.com wrote: Sane-backends is now in feature freeze. Bugs and documentation may still be fixed, but major improvements should be avoided. If you have access to a scanner or a platform that is even slightly uncommon, please try to test a current sane-backends git snapshot. FYI: Fedora 17, OS X 10.8 (Mountain Lion), Mingw32, and Mingw64 are still in good shape as of today's rebase. I'll do one more test close to Aug. 5th date. Chris
[sane-devel] Schedule for release of sane-backends 1.0.23
On Wed, Jul 18, 2012 at 6:44 PM, m. allan noah kitno455 at gmail.com wrote: Timetable: Jul 29, 2012: Feature freeze (only bugs + documentation updates) Aug 5, 2012: Code freeze (only horrible bugs + documentation updates) Aug 12, 2012: Release Speak up now if this timetable needs to change. In last month, I've done test compiles on Fedora 17, mingw32 (epson2 compiled only), mingw64 (epson2 only), and OS X 10.6 and we are good to go on those platforms. If I see new code submissions, I'll do another test compile around Aug 1 and fix any issues. Chris
[sane-devel] Windows pthread issue
On Wed, Feb 29, 2012 at 1:00 PM, Michael Cronenworth mike at cchtml.com wrote: Hi All/Chris, Fedora (17) is moving away from mingw.org to the w64 MinGW environment. The w64 folks have a new pthreads implementation called winpthreads and I have encountered an issue with it when compiling sane-backends. For the Windows build you try to expand the pthread_t pointer. Under winpthreads a pthread_t is not defined as a struct. Could you read the following support thread on the w64 sourceforge page and make any comments or corrections? https://sourceforge.net/tracker/index.php?func=detailaid=3495128group_id=202880atid=983355 Can you test a possible fix since you have access? They appear to be using an unsigned pointer so traditional -1 can't be used. In sanei_thread.c, modify these three blocks: #ifdef WIN32 pid-p = 0; [...] #ifdef WIN32 if (pid.p == 0) rc = SANE_TRUE; [...] #ifdef WIN32 rc = pid.p; to become: #ifdef WIN32 #ifdef WINTHREADS_API pid-p = 0; #else pid-p = NULL; #endif [...] #ifdef WIN32 #ifdef WINTHREADS_API if (pid == NULL) #else if (pid.p == 0) #endif rc = SANE_TRUE; [...] #ifdef WIN32 #ifdef WINTHREADS_API rc = (long)pid; #else rc = pid.p; #endif Chris Thanks, Michael
[sane-devel] [PATCH v2 03/14] bldchn, use our new macro SANE_CHECK_NETDB_FUNC
On Tue, Jan 24, 2012 at 1:49 PM, Ruediger Meier sweet_f_a at gmx.de wrote: On Tuesday 24 January 2012, Chris Bagwell wrote: On Mon, Jan 23, 2012 at 5:05 PM, Ruediger Meier sweet_f_a at gmx.de wrote: From: Ruediger Meier ruediger.meier at ga-group.nl Now we will find getaddrinfo and getnameinfo on win32 systems. Note it should be even safe to use that macro for all systems instead of trying AC_CHECK_FUNCS first. For now I don't want to do such risky things things and leave it as is. I would prefer if we went ahead and did your suggestion of doing check with new macro on all systems instead of this. ?I will be glad to test on a sampling of Unix systems to make sure there are no issues. So I'm saying I'd prefer this patch to change to delete old check for get* functions and do macro unconditionally. Ok, then I will polish SANE_CHECK_NETDB_FUNC a little for this: 1. maybe allowing multiple args 2. respecting cached results One thing I notice must be a bug. ?Right before the old call to AC_CHECK_FUNCS(), there is a check for BeOS and forces the variables to be no. ?Those need to be after the AC_CHECK_FUNCS to be meaningful. On first view I've also wondered about this but note that all ac_cv_* vars are cached variables and having them set to no before doing the check will result it this: checking for getaddrinfo... (cached) no checking for getnameinfo... (cached) no Oh yeah. Thanks for that info. I not so sure if the logic was written to work that way but its nice to know we were working. Chris
[sane-devel] [PATCH v2 03/14] bldchn, use our new macro SANE_CHECK_NETDB_FUNC
On Mon, Jan 23, 2012 at 5:05 PM, Ruediger Meier sweet_f_a at gmx.de wrote: From: Ruediger Meier ruediger.meier at ga-group.nl Now we will find getaddrinfo and getnameinfo on win32 systems. Note it should be even safe to use that macro for all systems instead of trying AC_CHECK_FUNCS first. For now I don't want to do such risky things things and leave it as is. I would prefer if we went ahead and did your suggestion of doing check with new macro on all systems instead of this. I will be glad to test on a sampling of Unix systems to make sure there are no issues. So I'm saying I'd prefer this patch to change to delete old check for get* functions and do macro unconditionally. One thing I notice must be a bug. Right before the old call to AC_CHECK_FUNCS(), there is a check for BeOS and forces the variables to be no. Those need to be after the AC_CHECK_FUNCS to be meaningful. I'm pretty sure BeOS support is broke in other ways. Anyways, I'll probably submit a patch to fix BeOS bug along with your patches. Chris --- ?configure ? ? ? ? ? ? ? ?| ?169 ++ ?configure.in ? ? ? ? ? ? | ? ?8 ++ ?include/sane/config.h.in | ? ?3 + ?3 files changed, 180 insertions(+), 0 deletions(-) diff --git a/configure b/configure index 4c9dfcd..5da8db8 100755 --- a/configure +++ b/configure @@ -9508,6 +9508,175 @@ fi +if test $ac_cv_func_getaddrinfo != yes \ + ?|| test $ac_cv_func_getnameinfo != yes; then + + + ?sane_netdb_includes=$ac_includes_default + ?#ifdef HAVE_SYS_SOCKET_H + ?# include sys/socket.h + ?#endif + ?#ifdef HAVE_NETDB_H + ?# include netdb.h + ?#endif + ?#ifdef HAVE_WINSOCK2_H + ?# include winsock2.h + ?#endif + ?#ifdef HAVE_WS2TCPIP_H + ?# include ws2tcpip.h + ?#endif + + ?ac_cv_func_getaddrinfo=no + ?need_winver_xp=no + + ?{ $as_echo $as_me:${as_lineno-$LINENO}: checking for win32 getaddrinfo 5 +$as_echo_n checking for win32 getaddrinfo... 6; } + ?cat confdefs.h - _ACEOF conftest.$ac_ext +/* end confdefs.h. ?*/ +$sane_netdb_includes +int +main () +{ +int testme(){ getaddrinfo; } + ?; + ?return 0; +} +_ACEOF +if ac_fn_c_try_compile $LINENO; then : + ?{ $as_echo $as_me:${as_lineno-$LINENO}: result: yes 5 +$as_echo yes 6; } + ? ?ac_cv_func_getaddrinfo=yes +else + ?{ $as_echo $as_me:${as_lineno-$LINENO}: result: no 5 +$as_echo no 6; } +fi +rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext + + ? ?if test $ac_cv_func_getaddrinfo != yes; then + ? ?{ $as_echo $as_me:${as_lineno-$LINENO}: checking for win32 getaddrinfo if WINVER is 0x0501 5 +$as_echo_n checking for win32 getaddrinfo if WINVER is 0x0501... 6; } + ? ?cat confdefs.h - _ACEOF conftest.$ac_ext +/* end confdefs.h. ?*/ +#define WINVER 0x0501 + ? ? ?$sane_netdb_includes +int +main () +{ +int testme(){ getaddrinfo; } + ?; + ?return 0; +} +_ACEOF +if ac_fn_c_try_compile $LINENO; then : + ?{ $as_echo $as_me:${as_lineno-$LINENO}: result: yes 5 +$as_echo yes 6; } + ? ? ?ac_cv_func_getaddrinfo=yes + ? ? ?need_winver_xp=yes +else + ?{ $as_echo $as_me:${as_lineno-$LINENO}: result: no 5 +$as_echo no 6; } +fi +rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext + ?fi + + ?if test $ac_cv_func_getaddrinfo = yes; then + +$as_echo #define HAVE_GETADDRINFO 1 confdefs.h + + + ? ?if test $need_winver_xp = yes; then + +$as_echo #define NEED_WINVER_XP 1 confdefs.h + + ? ?fi + ?fi + + ?unset sane_netdb_includes + + + ?sane_netdb_includes=$ac_includes_default + ?#ifdef HAVE_SYS_SOCKET_H + ?# include sys/socket.h + ?#endif + ?#ifdef HAVE_NETDB_H + ?# include netdb.h + ?#endif + ?#ifdef HAVE_WINSOCK2_H + ?# include winsock2.h + ?#endif + ?#ifdef HAVE_WS2TCPIP_H + ?# include ws2tcpip.h + ?#endif + + ?ac_cv_func_getnameinfo=no + ?need_winver_xp=no + + ?{ $as_echo $as_me:${as_lineno-$LINENO}: checking for win32 getnameinfo 5 +$as_echo_n checking for win32 getnameinfo... 6; } + ?cat confdefs.h - _ACEOF conftest.$ac_ext +/* end confdefs.h. ?*/ +$sane_netdb_includes +int +main () +{ +int testme(){ getnameinfo; } + ?; + ?return 0; +} +_ACEOF +if ac_fn_c_try_compile $LINENO; then : + ?{ $as_echo $as_me:${as_lineno-$LINENO}: result: yes 5 +$as_echo yes 6; } + ? ?ac_cv_func_getnameinfo=yes +else + ?{ $as_echo $as_me:${as_lineno-$LINENO}: result: no 5 +$as_echo no 6; } +fi +rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext + + ? ?if test $ac_cv_func_getnameinfo != yes; then + ? ?{ $as_echo $as_me:${as_lineno-$LINENO}: checking for win32 getnameinfo if WINVER is 0x0501 5 +$as_echo_n checking for win32 getnameinfo if WINVER is 0x0501... 6; } + ? ?cat confdefs.h - _ACEOF conftest.$ac_ext +/* end confdefs.h. ?*/ +#define WINVER 0x0501 + ? ? ?$sane_netdb_includes +int +main () +{ +int testme(){ getnameinfo; } + ?; + ?return 0; +} +_ACEOF +if ac_fn_c_try_compile $LINENO; then : + ?{ $as_echo $as_me:${as_lineno-$LINENO}: result: yes 5 +$as_echo
[sane-devel] [PATCH v2 04/14] bldchn, checking for socklen_t more safe
On Mon, Jan 23, 2012 at 5:05 PM, Ruediger Meier sweet_f_a at gmx.de wrote: From: Ruediger Meier ruediger.meier at ga-group.nl Here I just want to fix conflicts for win32 systems. Generally I'd say that we should never #define socklen_t int within config.h. But this may decide somebody else. Note this patch may break the build on systems where socklen_t is available now but the right headers are not included yet. For example for win32 you need to add ?#ifdef HAVE_WS2TCPIP_H ?#include ws2tcpip.h ?#endif wherever socklen_t is used. --- ?configure ? ?| ? 14 +++--- ?configure.in | ? 12 ++-- ?2 files changed, 21 insertions(+), 5 deletions(-) diff --git a/configure b/configure index 5da8db8..ba1df33 100755 --- a/configure +++ b/configure @@ -8257,12 +8257,20 @@ rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext ?{ $as_echo $as_me:${as_lineno-$LINENO}: result: $have_long_long 5 ?$as_echo $have_long_long 6; } -{ $as_echo $as_me:${as_lineno-$LINENO}: checking for socklen_t in sys/socket.h 5 -$as_echo_n checking for socklen_t in sys/socket.h... 6; } +{ $as_echo $as_me:${as_lineno-$LINENO}: checking for socklen_t 5 +$as_echo_n checking for socklen_t... 6; } ?cat confdefs.h - _ACEOF conftest.$ac_ext ?/* end confdefs.h. ?*/ -#include sys/socket.h +#ifdef HAVE_SYS_SOCKET_H +# include sys/socket.h +#endif +#ifdef HAVE_WINSOCK2_H +# include winsock2.h +#endif +#ifdef HAVE_WS2TCPIP_H +# include ws2tcpip.h +#endif ?int ?main () diff --git a/configure.in b/configure.in index 060b8b6..b30dd92 100644 --- a/configure.in +++ b/configure.in @@ -232,9 +232,17 @@ have_long_long=no ?AC_TRY_COMPILE([], [long long x; x = (long long)0;], AC_DEFINE(HAVE_LONG_LONG, 1, [Define if the long long type is available.]) have_long_long=yes) ?AC_MSG_RESULT($have_long_long) -AC_MSG_CHECKING([for socklen_t in sys/socket.h]) +AC_MSG_CHECKING([for socklen_t]) ?AC_TRY_COMPILE([ -#include sys/socket.h +#ifdef HAVE_SYS_SOCKET_H +# include sys/socket.h +#endif +#ifdef HAVE_WINSOCK2_H +# include winsock2.h +#endif +#ifdef HAVE_WS2TCPIP_H +# include ws2tcpip.h +#endif ?],[socklen_t len],AC_MSG_RESULT(yes), ?[AC_MSG_RESULT(no); AC_DEFINE(socklen_t,int, ?[Define socklen_t as \'int\' if necessary.])]) Since we are changing this, maybe we should change to more proper AC_CHECK_TYPES as well? That would require adding an #ifndef HAVE_SOCKELEN_T towards bottom of acinclude.m4 to match it. -- 1.7.6.1 -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org
[sane-devel] [PATCH v2 10/14] net, implement O_NONBLOCK stuff for win32
On Mon, Jan 23, 2012 at 5:05 PM, Ruediger Meier sweet_f_a at gmx.de wrote: From: Ruediger Meier ruediger.meier at ga-group.nl --- ?backend/net.c | ? ?5 + ?1 files changed, 5 insertions(+), 0 deletions(-) diff --git a/backend/net.c b/backend/net.c index 5946a1b..03133d3 100644 --- a/backend/net.c +++ b/backend/net.c @@ -2363,7 +2363,12 @@ sane_set_io_mode (SANE_Handle handle, SANE_Bool non_blocking) ? ? ? return SANE_STATUS_INVAL; ? ? } +#if ! defined _WIN32 ? if (fcntl (s-data, F_SETFL, non_blocking ? O_NONBLOCK : 0) 0) +#else + ?u_long iMode = non_blocking; + ?if (ioctlsocket(s-data, FIONBIO, iMode) != NO_ERROR) +#endif Is this UDP or TCP? If UDP, can you switch to sanei_udp_set_nonblock() that I recently added for epson2 network support? If TCP, can you add mirror function to sanie_tcp or maybe create a sanei_net.c for functions common to both UDP and TCP? Chris ? ? { ? ? ? DBG (1, sane_set_io_mode: fcntl failed (%s)\n, strerror (errno)); ? ? ? return SANE_STATUS_IO_ERROR; -- 1.7.6.1 -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org
[sane-devel] [PATCH v2 12/14] net, call WSAStartup on win32
On Mon, Jan 23, 2012 at 5:05 PM, Ruediger Meier sweet_f_a at gmx.de wrote: From: Ruediger Meier ruediger.meier at ga-group.nl We have to do this on win32 before doing winsock2 stuff. It affects the whole application so don't know wheather it's right to startup and clean it here within the net backend. Hmm, I wonder if the net.c backend can make more use of functions defined sanei_tcp.c? I've added the same logic there to open and close for use by TCP sockets in epson2 backend. Maybe the patch with net_read() and net_write() could make use of something there as well? Chris --- ?backend/net.c | ? 15 +++ ?1 files changed, 15 insertions(+), 0 deletions(-) diff --git a/backend/net.c b/backend/net.c index 0cd7f51..228bd59 100644 --- a/backend/net.c +++ b/backend/net.c @@ -1016,6 +1016,18 @@ sane_init (SANE_Int * version_code, SANE_Auth_Callback authorize) ? ? ? DBG (3, sane_init: Client has little endian byte order\n); ? ? } +#ifdef _WIN32 + ? ?{ + ? ? ?/* no much fun on win32 without WSAStartup but is this the right place? */ + ? ? ?WSADATA data; + ? ? ?int wsa_err = WSAStartup( MAKEWORD(2, 2), data) != 0; + ? ? ?if ( wsa_err != 0 ) + ? ? ? { + ? ? ? ? DBG (1, sane_init: WSAStartup failed with error %d\n, wsa_err ); + ? ? ? } + ? ?} +#endif + ?#ifndef NET_USES_AF_INDEP ? DBG (2, sane_init: determining sane service port\n); ? serv = getservbyname (sane-port, tcp); @@ -1196,6 +1208,9 @@ sane_exit (void) ? ? ? ?} ? ? ? free (devlist); ? ? } +#ifdef _WIN32 + ?WSACleanup(); +#endif ? DBG (3, sane_exit: finished.\n); ?} -- 1.7.6.1 -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org
[sane-devel] [PATCH v2 01/14] bldchn, provide macro SANE_CHECK_NETDB_FUNC
On Mon, Jan 23, 2012 at 5:05 PM, Ruediger Meier sweet_f_a at gmx.de wrote: From: Ruediger Meier ruediger.meier at ga-group.nl We want to use it later to do checks for getaddrinfo and getnameinfo more portable. --- ?acinclude.m4 | ? 57 + ?1 files changed, 57 insertions(+), 0 deletions(-) diff --git a/acinclude.m4 b/acinclude.m4 index d91c733..38d1819 100644 --- a/acinclude.m4 +++ b/acinclude.m4 @@ -486,6 +486,63 @@ AC_DEFUN([SANE_CHECK_GPHOTO2], ?]) ?# +# Checks for functions which comming usually from netdb.h or ws2tcpip.h (e.g. +# getaddrinfo). Header checks have to be done before as well as adding the right ... which usually come from ... Chris +# libs +AC_DEFUN([SANE_CHECK_NETDB_FUNC], +[ + ?sane_netdb_includes=$ac_includes_default + ?#ifdef HAVE_SYS_SOCKET_H + ?# include sys/socket.h + ?#endif + ?#ifdef HAVE_NETDB_H + ?# include netdb.h + ?#endif + ?#ifdef HAVE_WINSOCK2_H + ?# include winsock2.h + ?#endif + ?#ifdef HAVE_WS2TCPIP_H + ?# include ws2tcpip.h + ?#endif + + ?ac_cv_func_$1=no + ?need_winver_xp=no + + ?AC_MSG_CHECKING([for win32 ]$1) + ?AC_TRY_COMPILE( + ? ?[$sane_netdb_includes], + ? ?[int testme(){ ]$1[; }], + ? ?[AC_MSG_RESULT(yes) + ? ?ac_cv_func_$1=yes], + ? ?[AC_MSG_RESULT(no)]) + + ?dnl Under mingw, we may need to set WINVER to 0x0501 to expose the function. + ?if test $ac_cv_func_$1 != yes; then + ? ?AC_MSG_CHECKING([for win32 ]$1[ if WINVER is 0x0501]) + ? ?AC_TRY_COMPILE( + ? ? ?[#define WINVER 0x0501 + ? ? ?$sane_netdb_includes], + ? ? ?[int testme(){ ]$1[; }], + ? ? ?[AC_MSG_RESULT(yes) + ? ? ?ac_cv_func_$1=yes + ? ? ?need_winver_xp=yes], + ? ? ?[AC_MSG_RESULT(no)]) + ?fi + + ?if test $ac_cv_func_$1 = yes; then + ? ?AC_DEFINE([HAVE_]translit($1, [a-z], [A-Z]), [1], + ? ? ?[Define to 1 if you have the `]$1[' function.]) + + ? ?if test $need_winver_xp = yes; then + ? ? ?AC_DEFINE([NEED_WINVER_XP], [1], + ? ? ? ?[Define to 1 if some win32 functions needs WINVER set.]) + ? ?fi + ?fi + + ?unset sane_netdb_includes +]) + +# ?# Check for AF_INET6, determines whether or not to enable IPv6 support ?# Check for ss_family member in struct sockaddr_storage ?AC_DEFUN([SANE_CHECK_IPV6], -- 1.7.6.1 -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org
[sane-devel] [PATCH 00/13] win32 portable net backend
Ruediger, I'm try to find time and comment on this soon. Should I comment on github or should you post the patches to the list so I can reply? I had at least 3 comments from quick glance. Also, you might should request write access to git so you can commit these patches yourself if you think you'll be cleaning up for a while; although I have no problems doing that on your behalf. Even if you commit yourself, I think your current process is good though since your touching framework and other people's backends. If its more than a trivial #ifdef HAVE_HEADER_FILE addition, I also post framwork/other's backend patches to list before committing. Chris On Sun, Jan 22, 2012 at 1:35 AM, Ruediger Meier sweet_f_a at gmx.de wrote: Hi, I've spend some time making the net backend win32 portable. Maybe we could build a useful native (non-cygwin) xsane (saned client) on Windows now. Please pull from my private clone git://github.com/rudimeier/sane-backends.git branch win32/net Hope this doesn't break something. If wanted I would post these patches on this list too. Ruediger Meier (13): ?bldchn, provide macro SANE_CHECK_NETDB_FUNC ?bldchn, add checks for netdb.h and ws2tcpip.h ?bldchn, use our new macro SANE_CHECK_NETDB_FUNC ?bldchn, checking for socklen_t more safe ?fix sanei_udp, socklen_t was not defined on win32 ?bldchn, don't always disable net backend on win32 ?net, move some includes from net.h to net.c ?net, add missing includes for win32 ?net, re-use sane_set_io_mode ?net, implement O_NONBLOCK stuff for win32 ?net, use recv/send instead of read/write ?net, call WSAStartup on win32 ?bldchn, SANE_CHECK_IPV6 more portable ?acinclude.m4 ? ? ? ? ? ? | ? 76 +++--- ?backend/net.c ? ? ? ? ? ?| ? 54 +++- ?backend/net.h ? ? ? ? ? ?| ? ?3 - ?configure ? ? ? ? ? ? ? ?| ?201 ++ ?configure.in ? ? ? ? ? ? | ? 22 +- ?include/sane/config.h.in | ? ?9 ++ ?sanei/sanei_udp.c ? ? ? ?| ? ?3 + ?7 files changed, 329 insertions(+), 39 deletions(-) -- 1.7.6.1 -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org
[sane-devel] Subject: [PATCH 0/4] fix and cleanup portable PATH_SEP and DIR_PATH defines
On Tue, Jan 17, 2012 at 6:13 PM, Ruediger Meier sweet_f_a at gmx.de wrote: On Wednesday 18 January 2012, Chris Bagwell wrote: On Tue, Jan 17, 2012 at 3:01 PM, Ruediger Meier wrote: Thx, I just see I've messed up that patch series a little while playing around with my MUA. Could you cleanly apply it? If not then I would reroll it. I haven't tried yet. ?Can you resend directly to me to save time? You can send as attachements to me until you get MUA worked out. Please pull from my private clone git://github.com/rudimeier/sane-backends.git branches win32/path win32/scsi Thanks. That was easiest way. I've just pushed your changes. I added a change log entry to your second patch series. I appreciate the scsi fix. I was struggle with how to fix that since it was reported not compiling with mingw64. Chris
[sane-devel] Subject: [PATCH 0/4] fix and cleanup portable PATH_SEP and DIR_PATH defines
This patch series makes minor portability fixes to following backends. I've not seen any objections from various maintainers... so if I do not hear back by weekend, I'll commit this patch series. canon_dr cardscan epjitsu fujitsu kodak gt68xx microtek2 dll Thanks for this and other windows patches, Ruediger. Chris On Sun, Jan 15, 2012 at 8:31 AM, Ruediger Meier sweet_f_a at gmx.de wrote: From: Ruediger Meier ruediger.meier at ga-group.nl This patch-set is to fix and unify PATH_SEP and DIR_PATH defines in the context of seeing them conflicting on cygwin. Ruediger Meier (4): ?fix, PATH_SEP and DIR_SEP were inconsistently defined on cygwin since ? ? ? ?5168206e (also see a28bd843) ?cleanup, remove all unused PATH_SEP defines ?cleanup, always use the same portable PATH_SEP define (note _WIN32 ? ? is not defined on cygwin) ?update ChangeLog for previous commits ?ChangeLog ? ? ? ? ? ?| ? ?6 ++ ?backend/canon_dr.h ? | ? ?8 ?backend/cardscan.h ? | ? ?8 ?backend/dll.c ? ? ? ?| ? ?6 +- ?backend/epjitsu.h ? ?| ? ?8 ?backend/fujitsu.h ? ?| ? ?8 ?backend/gt68xx.c ? ? | ? ?2 +- ?backend/kodak.h ? ? ?| ? ?8 ?backend/microtek2.h ?| ? ?2 +- ?sanei/sanei_access.c | ? ?2 +- ?sanei/sanei_config.c | ? ?5 + ?11 files changed, 11 insertions(+), 52 deletions(-) -- 1.7.6.1 -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org
[sane-devel] Subject: [PATCH 0/4] fix and cleanup portable PATH_SEP and DIR_PATH defines
On Tue, Jan 17, 2012 at 3:01 PM, Ruediger Meier sweet_f_a at gmx.de wrote: On Tuesday 17 January 2012, Chris Bagwell wrote: This patch series makes minor portability fixes to following backends. I've not seen any objections from various maintainers... so if I do not hear back by weekend, I'll commit this patch series. Thx, I just see I've messed up that patch series a little while playing around with my MUA. Could you cleanly apply it? If not then I would reroll it. I haven't tried yet. Can you resend directly to me to save time? You can send as attachements to me until you get MUA worked out. BTW is this list the right one to send patches at all? Yes. Although most active developers have write access and commit themselves. So you don't always see a lot of patches to this list. Chris
[sane-devel] MinGW w64 build results
On Tue, Dec 20, 2011 at 1:53 PM, Michael Cronenworth mike at cchtml.com wrote: Chris Bagwell on 12/18/2011 09:56 PM wrote: Not really. ?I'm sure upstream libusb is the best place to discuss. As long as libusb project has a suggested work around/file location for dealing with w64, we can accommodate it as well as standard usb.h easy enough. After thinking about it, I'm not sure libusb is interested in Windows. libusb-win32[1] though, is, and it is the project I use for libusb support. They renamed their include file usb.h to lusb0_usb.h in their latest release. Unfortunately they are not using MinGW (and state they have no plans for it) to build from source. Since the original libusb project will never compile/run on Windows I think a define should be added to use the libusb-win32 include file name. Even if you can't compile everything using MinGW. [1] https://sourceforge.net/projects/libusb-win32/ I had some time to come up to speed with this and added support for libusb-win32 to sane-backends. I was able to compile and link the epson2 backend under mingw32 and using the libusb.a and lusb0_usb.h that libusb-win32 distributes. Chris
[sane-devel] [iscan] New release available
Pushed your patch. Chris On Thu, Dec 8, 2011 at 5:58 PM, Olaf Meeuwissen olaf.meeuwissen at avasys.jp wrote: Dear all, A new release of Image Scan! for Linux has been made available at our website[1]. ?This release updates both the core and data packages. ?For information on which versions of what packages you need, please see the FAQ[2] on that topic. ?[1] http://avasys.jp/eng/linux_driver/ ?[2] http://avasys.jp/eng/linux_driver/faq/id000651.php Changes to the epkowa.desc file against the latest version in the git repository have been attached. ?If someone could yet again commit these to sane-backends that would be appreciated. Below is a summary of the changes since the last announcement on this mailing list. ?-- ?iscan-2.28.1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?(2011-11-18) ?-- ?* the quick-format option has been deprecated and will be removed in ? ?a future release. ?Use the scan-area option instead. ?* fixes hangs when scanning from the GIMP ?* fixes shifting of the selected scan area when scanning with the ADF ? ?unit on some models ?* fixes handling of some predefined scan area sizes ?* fixes an issue when used in combination with saned Noteworthy changes to the data package: ?-- ?iscan-data-1.13.0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? (2011-11-15) ?-- ?* adds support for the following devices: ? ?- Epson ME OFFICE 535 ? ?- Epson ME OFFICE 940FW Series ? ?- Epson Stylus NX635 ? ?- Epson Stylus Office BX535WD ? ?- Epson Stylus Office BX630FW Series ? ?- Epson Stylus Office BX935FWD ? ?- Epson Stylus SX535WD ? ?- K300 Series ? ?- PX-1700F ? ?- WP-4511 ? ?- WP-4521 ? ?- WorkForce 645 ? ?- WorkForce 845 ? ?- WorkForce K301 Hope this helps, -- Olaf Meeuwissen, LPIC-2 ? ? ? ? ? FLOSS Engineer -- AVASYS CORPORATION FSF Associate Member #1962 ? ? ? ? ? ? ? Help support software freedom ? ? ? ? ? ? ? ? http://www.fsf.org/jf?referrer=1962 -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org
[sane-devel] MinGW w64 build results
On Fri, Dec 16, 2011 at 2:22 PM, Michael Cronenworth mike at cchtml.com wrote: MinGW thread time again. The MinGW world has been split into two trees. The original mingw.org tree and the new w64 tree. Fedora, and other distros, are migrating from mingw.org to w64. Sane-backends will compile and run under the w64 environment with a few caveats. 1. The DDK includes are not supposed to be directly used. ? sanei/sanei_scsi.c and tools/sane-find-scanner.c needed the attached patches. I looked at this change but as it is, it breaks compiles on Fedora 16. It looks like including the DDK files is the only option there. Is there a standard way to check when your on a w64 platform? I guess we could move some header file checks to configure to detect were to get SCSI symbols. 2. w64 includes additional Windows API headers, which now include usb.h and this conflicts with the libusb header of the same name. This prevents sane from including USB support as sane looks for libusb and not the Windows API. For now I renamed the w64 header and used the libusb header and the resulting libsane.dll works with my USB scanner in Windows. I have not contacted libusb yet, but I feel they need to move their header into a libusb/ sub-dir, but this will of course break lots of apps. Any suggestions? Not really. I'm sure upstream libusb is the best place to discuss. As long as libusb project has a suggested work around/file location for dealing with w64, we can accommodate it as well as standard usb.h easy enough. Chris Thanks, Michael -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org
[sane-devel] TPU functionality in epson2-backend.
I just saw this commit so looks like its in now. The following commit has been merged in the master branch: commit 56cac490f968008084cb88604ec633659bd36dae Author: Alessandro Zummo a.zummo at towertech.it Date: Fri Dec 2 00:27:28 2011 +0100 epson2: support TPU2 on GT-X800/V700 On Thu, Dec 1, 2011 at 5:41 AM, Martin Zackrisson martin.zackrisson at cmb.gu.se wrote: Just for reference, I now have the programming guide for the scanner and it seems that the capability that I'm looking to implement in fact is structured as a second TPU-mode and not a setting in the current TPU mode. So I'll get back to scratch and see if I can push it into TPU 2 instead. Solved! The following changes is ALL that is required for the V700 (or GT-X900 as it also is called) to implement the full TPU area. Is there a possibility of getting that into the official epson2 backend? epson2-ops.c: 673c673 ? ? ? ? if (e2_model(s, GT-X800) || e2_model(s, GT-X900)) { --- ? ? ? ? if (e2_model(s, GT-X800)) { 1223c1223 ? ? if (s-lcount 3 (e2_model(s, GT-X800) || e2_model(s, GT-X900))) { --- ? ? if (s-lcount 3 e2_model(s, GT-X800)) { /Martin -- Martin Zackrisson G?teborgs Universitet / Cell ?Molekyl?rbiologi Medecinaregatan 9e SE-413 90 G?teborg Sweden Phone +46 (0)31 786 3753 -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ?to sane-devel-request at lists.alioth.debian.org
[sane-devel] [iscan] New release available
On Thu, Nov 10, 2011 at 10:02 PM, Olaf Meeuwissen olaf.meeuwissen at avasys.jp wrote: Olaf Meeuwissen olaf.meeuwissen at avasys.jp writes: Dear all, A new release of Image Scan! for Linux has been made available at our website[1]. ?This release updates both the core and data packages. ?For information on which versions of what packages you need, please see the FAQ[2] on that topic. ?[1] http://avasys.jp/eng/linux_driver/ ?[2] http://avasys.jp/eng/linux_driver/faq/id000651.php Changes to the epkowa.desc file against the latest version in the git repository have been attached. ?If someone could yet again commit these to sane-backends that would be appreciated. OK. Finally had a minute to upload this. Chris
[sane-devel] Fwd: [sane-commit] [SCM] SANE backends - scanner drivers branch, master, updated. RELEASE_1_0_22-179-g1e4595c
Yiks. How did I get a u as only a 1000 in my head. The test scans worked so maybe the original values were a little high. Anyways, this means I also messed up a patch for umax-pp. I'll go the opposite direction since it proved to be confusing at least once to me. I'll create a posix compatible sleep() using usleep for platforms that do not have it and then people can more comfortably think in seconds when needed. Then I'll revert all my sleep-usleep changes. Thanks much! Chris On Wed, Nov 9, 2011 at 10:47 PM, Stef stef.dev at free.fr wrote: ? ? ? ?Hello, ? ? ? ?since usleep() time value is expressed in microseconds, I think ? ? ? ?sleep(5) should be replaced by usleep(500UL). usleep(5000) is too short. Regards, ? ? ? ?Stef -- Forwarded message -- From:?Chris Bagwell chris at cnpbagwell.com To:?sane-commit at lists.alioth.debian.org Date:?Thu, 10 Nov 2011 04:00:08 + Subject:?[sane-commit] [SCM] SANE backends - scanner drivers branch, master, updated. RELEASE_1_0_22-179-g1e4595c The following commit has been merged in the master branch: commit 50475719ef40038b4c73e13dc164d72b9995527e Author: Chris Bagwell chris at cnpbagwell.com Date: ? Wed Nov 9 21:50:18 2011 -0600 ? ?portabilty mods to epson2 backend ? ?No behavior changes. ?Wrap some optional headers with #ifdef's and ? ?use usleep() instead of sleep() since it exist everywhere. diff --git a/backend/epson2-ops.c b/backend/epson2-ops.c index 575340c..e53f33c 100644 --- a/backend/epson2-ops.c +++ b/backend/epson2-ops.c @@ -20,9 +20,9 @@ ?#include sane/config.h ?#include unistd.h ? ? ? ? ? ?/* sleep */ - +#ifdef HAVE_SYS_SELECT_H ?#include sys/select.h - +#endif ?#include byteorder.h @@ -1439,7 +1439,7 @@ e2_wait_button(Epson_Scanner * s) ? ? ? ? ? ? ? ? ? ? ? ?if (button_status) ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?s-hw-wait_for_button = SANE_FALSE; ? ? ? ? ? ? ? ? ? ? ? ?else - ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? sleep(1); + ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? usleep(1000); ? ? ? ? ? ? ? ?} else { ? ? ? ? ? ? ? ? ? ? ? ?/* we run into an error condition, just continue */ ? ? ? ? ? ? ? ? ? ? ? ?s-hw-wait_for_button = SANE_FALSE; @@ -1533,7 +1533,7 @@ e2_wait_warm_up(Epson_Scanner * s) ? ? ? ? ? ? ? ? ? ? ? ? ? ?s-retry_count); ? ? ? ? ? ? ? ? ? ? ? ?return SANE_STATUS_DEVICE_BUSY; ? ? ? ? ? ? ? ?} - ? ? ? ? ? ? ? sleep(5); + ? ? ? ? ? ? ? usleep(5000); ? ? ? ?} ? ? ? ?return SANE_STATUS_GOOD; @@ -1816,7 +1816,7 @@ read_info_block(Epson_Scanner * s, EpsonDataRec * result) ? ? ? ? ? ? ? ?if (ext_status[0] EXT_STATUS_WU) { ? ? ? ? ? ? ? ? ? ? ? ?free(ext_status); - ? ? ? ? ? ? ? ? ? ? ? sleep(5); ? ? ? /* for the next attempt */ + ? ? ? ? ? ? ? ? ? ? ? usleep(5000); ? /* for the next attempt */ ? ? ? ? ? ? ? ? ? ? ? ?DBG(1, retrying ESC G - %d\n, ++(s-retry_count)); diff --git a/backend/epson2.c b/backend/epson2.c index bdd9c27..a07510f 100644 --- a/backend/epson2.c +++ b/backend/epson2.c @@ -389,8 +389,6 @@ e2_network_discovery(void) ? ? ? ?struct timeval to; - ? ? ? long save_flags, flags; - ? ? ? ?status = sanei_udp_open_broadcast(fd); ? ? ? ?if (status != SANE_STATUS_GOOD) ? ? ? ? ? ? ? ?return; @@ -405,9 +403,7 @@ e2_network_discovery(void) ? ? ? ?FD_ZERO(rfds); ? ? ? ?FD_SET(fd, rfds); - ? ? ? save_flags = flags = fcntl(fd, F_GETFL, 0L); - ? ? ? flags |= O_NONBLOCK; - ? ? ? fcntl(fd, F_SETFL, flags); + ? ? ? sanei_udp_set_nonblock(fd, SANE_TRUE); ? ? ? ?if (select(fd + 1, rfds, NULL, NULL, to) 0) { ? ? ? ? ? ? ? ?while ((len = sanei_udp_recvfrom(fd, buf, 76, ip)) == 76) { ? ? ? ? ? ? ? ? ? ? ? ?DBG(5, response from %s\n, ip); @@ -417,7 +413,6 @@ e2_network_discovery(void) ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?attach_one_net(ip); ? ? ? ? ? ? ? ?} ? ? ? ?} - ? ? ? fcntl(fd, F_SETFL, save_flags); ? ? ? ?DBG(5, %s, end\n, __func__); @@ -650,7 +645,7 @@ device_detect(const char *name, int type, SANE_Status *status) ? ? ? ? ? ? ? ? ? ? ? ? * sleep a bit. ? ? ? ? ? ? ? ? ? ? ? ? */ ? ? ? ? ? ? ? ? ? ? ? ?if (dev-connection == SANE_EPSON_NET) - ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? sleep(1); + ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? usleep(1000); ? ? ? ? ? ? ? ? ? ? ? ?return scanner_create(dev, status); ? ? ? ? ? ? ? ?} diff --git a/backend/epson2.h b/backend/epson2.h index b448784..807251a 100644 --- a/backend/epson2.h +++ b/backend/epson2.h @@ -22,7 +22,9 @@ ?#define BACKEND_NAME epson2 ?#define DEBUG_NOT_STATIC +#ifdef HAVE_SYS_IOCTL_H ?#include sys/ioctl.h +#endif ?#ifdef HAVE_STDDEF_H ?#include stddef.h -- SANE backends - scanner drivers ___ sane-commit mailing list sane-commit at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-commit -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password
[sane-devel] MinGW status
Hi all, I have 1 last patch to submit (the pthreads I posted to list) and all of Sane's framework will compile under MinGW. At least USB and Network parts of framework have been verified to work. I know that the Epson2 and Fujitsu backends compile and work under windows with resulting executable. Most other backends will not compile and must be disabled using configure BACKENDS=epson2 fujitsu. The compile failures mostly look like minor issues to fix but I'll leave that for another time. I don't want to chance the epson2 usleep type issue again on backends I can't test. I will fix a specific backend on request (just send me an email). I can either submit the fix or send you the diff for you to review and submit. Chris
[sane-devel] Fwd: [sane-commit] [SCM] SANE backends - scanner drivers branch, master, updated. RELEASE_1_0_22-173-gc64249f
Your right. I'll fix it. When I noticed I had killed Linux compiles in git, I rushed to fast to get the fix in and restored to much logic from the original code it was based on in espon2.c backend. Thank you much for reviewing the diff! Chris On Wed, Nov 9, 2011 at 8:31 AM, Stef stef.dev at free.fr wrote: ? ? ? ?Hello, ? ? ? ?I don't get what save_flags var is for: ? ? ? ?save_flags = flags = fcntl(fd, F_GETFL, 0L); ? ? ? ?it is not used anywhere in sanei_udp_set_nonblock(). Regards, ? ? ? ?Stef -- Forwarded message -- From:?Chris Bagwell chris at cnpbagwell.com To:?sane-commit at lists.alioth.debian.org Date:?Wed, 09 Nov 2011 03:23:34 + Subject:?[sane-commit] [SCM] SANE backends - scanner drivers branch, master, updated. RELEASE_1_0_22-173-gc64249f The following commit has been merged in the master branch: commit c64249fc318aaef44f8be89f3f4480553904c7eb Author: Chris Bagwell chris at cnpbagwell.com Date: ? Tue Nov 8 21:22:42 2011 -0600 ? ?fix compile failure on non-windows platform. diff --git a/sanei/sanei_udp.c b/sanei/sanei_udp.c index 3e84761..795b2de 100644 --- a/sanei/sanei_udp.c +++ b/sanei/sanei_udp.c @@ -45,6 +45,7 @@ ?#include unistd.h ?#include stdlib.h ?#include string.h +#include fcntl.h ?#ifdef HAVE_WINSOCK2_H ?#include winsock2.h @@ -194,7 +195,7 @@ sanei_udp_set_nonblock(int fd, SANE_Bool nonblock) ? ? ? ?ioctlsocket(fd, FIONBIO, mode); ?#else - ? ? ? long flags; + ? ? ? long save_flags, flags; ? ? ? ?save_flags = flags = fcntl(fd, F_GETFL, 0L); ? ? ? ?if (nonblock) -- SANE backends - scanner drivers ___ sane-commit mailing list sane-commit at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-commit -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org
[sane-devel] 64-bit compile (was:Thread code review and testers)
On Mon, Nov 7, 2011 at 11:27 AM, Michael Cronenworth mike at cchtml.com wrote: Michael Cronenworth wrote: The 64-bit compile against the latest git code doesn't complete. The backend make doesn't include sanei_magic.lo so the fujitsu object has some undefined function calls. It seems this is a byproduct of using PRELOADABLE_BACKENDS. I've attached a patch for it. Thanks for patch. I'll commit it soon.
[sane-devel] Thread code review and testers
On Mon, Nov 7, 2011 at 10:55 AM, Michael Cronenworth mike at cchtml.com wrote: Michael Cronenworth wrote: I'll try the latest git checkout plus this patch and see if it works any better for me. The latest git + only your patch now compiles on the first try. However, the resulting DLL doesn't find the scanner. sane-find-scanner.exe finds the USB device, but scanimage.exe and my program that uses libsane.dll do not find the scanner. I had a feeling it is looking for the config file to reference the USB vendor ID and product ID, but it cannot find it. I moved the *.conf files into the same directory as the runtime DLL and EXE files and now the scanner is found. Scanning works, too. Glad to hear its working! Your doing better then me so far... but I'm almost there. So far, I can't get the preloaded backends to kick in from libsane-1.dll. Did you do anything special to get it working? I just committed a fix to libtool that lets us now correctly build each individual backend as its own DLL. I'm not sure I'll get the urge to fix this soon but it wouldn't be hard to get the dll backend to load these DLL's now that we are correctly building them. After that DLL build fix, I can copy libsane-epson2-1.dll to same directory as scanimage.exe but name it libsane-1.dll and now the epson2 backend is working with scanimage.exe. It correctly detects my network epson scanner but then fails to open the socket correctly. Hopefully, it will be a simple fix. I would guess that between .21 and .22 that the config file handling changed? It would be nice to have the Windows binaries look for the config files in a few sub-directories besides the current directory. (./etc, ../etc ./) The logic was modified but don't see exactly any difference in behaviour. The best solution right now is to set SANE_CONFIG_DIR to where ever you store them. Maybe in future we can modify to use standard Windows DLL search logic and look inside same directory as .exe. I've never done that before but I guess it wouldn't be hard. Unfortunately this trick doesn't make the 64-bit DLL work. Something else is wrong. I'm glad to hear in your other post that 64-bit is now working for you as well! I'm sure there are tons of people that would like to run a native Windows executable (non-cygwin) from command line/background task to do some scanning. Chris
[sane-devel] Thread code review and testers
On Mon, Nov 7, 2011 at 7:44 PM, Chris Bagwell chris at cnpbagwell.com wrote: After that DLL build fix, I can copy libsane-epson2-1.dll to same directory as scanimage.exe but name it libsane-1.dll and now the epson2 backend is working with scanimage.exe. It correctly detects my network epson scanner but then fails to open the socket correctly. ?Hopefully, it will be a simple fix. I got it working. It was minor changes to enable winsock2 correctly and how to set up non-blocking. Now I can do network scans from my epson scanner using the epson2 backend. On top of that, I'm able to compile the scanimage.exe from my Fedora box and I'm running the scanimage.exe using wine from same box. Now thats the way to develop windows applications. :-) Chris
[sane-devel] Thread code review and testers
Hi all, I'm looking for a reviewer of attached patch and I'm also hoping someone has a scanner who's driver uses threads can test this patch to verify no regressions occurred (be sure and use --enable-pthread on configure). Current code assumes/requires that SANE_Pid is represented as an integer on all platforms because we do pid == -1 checks everywhere. Pthreads says that it can not be assumed pthread_t is an integer and on windows platform it in fact is not an integer. Attached is my solution to abstract out checks for valid PID's so that it will work with windows and any other pthread implementation that is similar. For each backend that gets ported to mingw and uses threads, it will need to switch to use sanei_thread_is_invalid(pid) instead of (pid == -1). That was not done as a part of this test patch. Chris -- next part -- A non-text attachment was scrubbed... Name: pthread.diff Type: text/x-patch Size: 4374 bytes Desc: not available URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/2006/e7b07e80/attachment.bin
[sane-devel] Windows scanner issue
On Fri, Nov 4, 2011 at 12:38 PM, Michael Cronenworth mike at cchtml.com wrote: Hello, In the past I posted about building for Windows. I had SANE 1.0.21 working fine. Today I have tried to build 1.0.22 and test it. The build works fine, but when I use the resulting DLL the scanner cannot be seen. Switching back to the 1.0.21 DLL makes the scanner visible again. Is there anything that changed between 21 and 22 that could affect this? I can't comment much on this part. Lots of changes occurred. Thanks, Michael P.S. It seems my Win32 patches have collected dust and are still not in SANE. Would anyone be willing to work with me to have them added? I've had your patches sitting in my inbox since 2009. Never got around to it. Sorry. Since that time, I finally installed the mingw32 cross compiler that Fedora has so now its easier for me to do this type stuff. I'll start peeling off some of the fixes and submit them. The main thing preventing me from doing it earlier is about 1/2 are generic issues to be solved and so I'm working on ways not to use the #ifdef WIN32 check. Lets hope I don't get distracted again. :-) Chris
[sane-devel] How to change the Makefile.in and Makefile.am
I'll also add you should be sure and look at doc/backend-writing.txt for tips on what to add to Makefile.am. If there is info that would have helped you further, patches to that file would be ideal to submit for the next developers benefit. For example, the tip about only needing to sending patches for Makefile.am and configure.in might be good to add. Chris On Thu, Sep 15, 2011 at 7:44 AM, m. allan noah kitno455 at gmail.com wrote: You must add the backend name to configure.in, and various backend file names in several places in backend/Makefile.am. Look for an existing backend name to see all the places. Do the same for your manpages and description files in doc/. Then run automake (rebuilds all the Makefile.in files) and then run autoconf to rebuild configure. There are probably other steps I forgot ;) If you send a patch to the list, only send the configure.in and Makefile.am changes, I'll re-run automake and autoconf here. allan On Thu, Sep 15, 2011 at 2:20 AM, ?kingjian_tao at avision.com.cn wrote: Dear, ? ?I write a new backend,I success add it to sane-backends-1.0.18,But ?now I want to add this backend to sane-backends-1.0.22. Could you tell me which change should I do.and How to change the Makefile.in and Makefile.am,Thanks. Best Regards -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org -- The truth is an offense, but not a sin -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org
[sane-devel] autoconf/configure and debug
On Tue, Jun 14, 2011 at 2:49 PM, stef stef.dev at free.fr wrote: ? ? ? ?Hello, ? ? ? ?in order to debug more efficiently, I have to re-edit Makefiles to remove the two -O2 options in CFLAGS and change the -g to the -ggdb option. Is there an easy way to get CFLAGS=-ggdb when configuring SANE ? I've not tried this with sane-backends but its supposed to be as simple as: ./configure CFLAGS=-ggdb configure will not default to CFLAGS=-g -O2 if you specify your own value. But if your getting two -O2 then something else may be adding it as well. Chris
[sane-devel] sane backend build system - help
Have you looked at doc/backend-writing.txt? I suggest taking notes as you work threw this process and provide patches to that file to improve it for next author. On Mon, Apr 25, 2011 at 9:05 AM, Paul Newall p.newalls at ntlworld.com wrote: Progress: I managed to copy an existing backend, and add it to the build system. So now I can start the actual backend writing process. That was quite hard work. I'm not sure if that was just me, or whether it might encourage other people to have a go at writing a backend if there was an easier way to get started. Maybe some experienced backend writer could provide one of the following? A build and install system for a single backend. I think that would be the easiest to understand. And the install process would be faster during developement if you were only installing one backend. It would need to be easy to change the backend name, perhaps by changing one variable with the name in? Or, provide one empty backend as part of the sane-backends, that you can experiment with without having to copy and rename an existing backend. Also it would be nice to have empty backend files: emptybackend.c emptybackend.h emptybackend.conf with lots of comments in about what you should add in various places. - Original Message - From: Paul Newall p.newalls at ntlworld.com To: sane-devel at lists.alioth.debian.org Sent: Sunday, April 24, 2011 8:26 PM Subject: Re: [sane-devel] sane backend build system - help autoreconf -i -f seems to have enabled me to make the backends. So now I am trying the approach of adding a new backend (copy of an existing one) to the backends to see if I can install that. Paul - Original Message - From: Chris Bagwell chris at cnpbagwell.com To: Paul Newall p.newalls at ntlworld.com Cc: sane-devel at lists.alioth.debian.org Sent: Sunday, April 24, 2011 7:15 PM Subject: Re: [sane-devel] sane backend build system - help Its common to have a newer version of autofoo tools on your system and you'll need to update the pre-existing files committed in git. Try running with install option to get newer versions of files: autoreconf -i If that doesn't work then use the force option: ?autoreconf -i -f Chris On Sun, Apr 24, 2011 at 10:17 AM, Paul Newall p.newalls at ntlworld.com wrote: I want to write a new backend. Currently the main obstacle is that I find the backend build system very intimidating. I have no experience of autoconf, automake, or libtool. On my system (ubuntu 10.10) I can run autoreconf and ./configure apparently OK, but when I run make I get these errors: libtool: Version mismatch error. This is libtool 2.2.10, but the libtool: definition of this LT_INIT comes from libtool 2.2.6b. libtool: You should recreate aclocal.m4 with macros from libtool 2.2.10 libtool: and run autoconf again. make[2]: *** [libsane_la-dll-s.lo] Error 63 How would I recreate aclocal.m4 with macros from libtool 2.2.10 ? I don't have much idea what that means. Is it best to start by adding an empty new backend to the build system? (in which case I will need to get the build system to work first) Or would it be better to make and install the new backend independently during development? (this might be easier? and it might give me a faster modify, build, test cycle during development? But are there difficulties that I have not discovered yet?) -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ?to sane-devel-request at lists.alioth.debian.org -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ?to sane-devel-request at lists.alioth.debian.org
[sane-devel] sane backend build system - help
Its common to have a newer version of autofoo tools on your system and you'll need to update the pre-existing files committed in git. Try running with install option to get newer versions of files: autoreconf -i If that doesn't work then use the force option: autoreconf -i -f Chris On Sun, Apr 24, 2011 at 10:17 AM, Paul Newall p.newalls at ntlworld.com wrote: I want to write a new backend. Currently the main obstacle is that I find the backend build system very intimidating. I have no experience of autoconf, automake, or libtool. On my system (ubuntu 10.10) I can run autoreconf and ./configure apparently OK, but when I run make I get these errors: libtool: Version mismatch error.? This is libtool 2.2.10, but the libtool: definition of this LT_INIT comes from libtool 2.2.6b. libtool: You should recreate aclocal.m4 with macros from libtool 2.2.10 libtool: and run autoconf again. make[2]: *** [libsane_la-dll-s.lo] Error 63 How would I recreate aclocal.m4 with macros from libtool 2.2.10 ? I don't have much idea what that means. Is it best to start by adding an empty new backend to the build system? (in which case I will need to get the build system to work first) Or would it be better to?make and install ?the new backend independently during development? (this might be easier? and it might give me a faster modify, build, test cycle during development? But are there difficulties that I have not discovered yet?) -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org
[sane-devel] message in ./configure chatter
Thanks for taking time to report. This warning message can be ignored in this case. At least its not the source of issue your seeing. Chris On Sat, Mar 19, 2011 at 11:44 PM, Edward Hartouni ed_hartouni at comcast.net wrote: on my iMac running OSX 10.5.8 with the sane-backends-1.0.22 ./configure script found the message below. the binaries are not working for my Nikon Coolscan 4000 backend coolscan3 (no scan takes place, though it is possible to control some of the functions on the scanner)... checking IOKit/scsi/SCSITaskLib.h usability... no checking IOKit/scsi/SCSITaskLib.h presence... yes configure: WARNING: IOKit/scsi/SCSITaskLib.h: present but cannot be compiled configure: WARNING: IOKit/scsi/SCSITaskLib.h: ? ? check for missing prerequisite headers? configure: WARNING: IOKit/scsi/SCSITaskLib.h: see the Autoconf documentation configure: WARNING: IOKit/scsi/SCSITaskLib.h: ? ? section Present But Cannot Be Compiled configure: WARNING: IOKit/scsi/SCSITaskLib.h: proceeding with the compiler's result configure: WARNING: ? ? ## - ## configure: WARNING: ? ? ## Report this to?sane-devel at lists.alioth.debian.org?## configure: WARNING: ? ? ## - ## checking for IOKit/scsi/SCSITaskLib.h... no checking IOKit/cdb/IOSCSILib.h usability... no checking IOKit/cdb/IOSCSILib.h presence... no checking for IOKit/cdb/IOSCSILib.h... no checking IOKit/scsi/SCSICommandOperationCodes.h usability... yes checking IOKit/scsi/SCSICommandOperationCodes.h presence... yes checking for IOKit/scsi/SCSICommandOperationCodes.h... yes checking IOKit/scsi-commands/SCSICommandOperationCodes.h usability... yes checking IOKit/scsi-commands/SCSICommandOperationCodes.h presence... yes checking for IOKit/scsi-commands/SCSICommandOperationCodes.h... yes checking scsi.h usability... no checking scsi.h presence... no checking for scsi.h... no checking sys/scsi.h usability... no checking sys/scsi.h presence... no checking for sys/scsi.h... no checking sys/scsicmd.h usability... no checking sys/scsicmd.h presence... no checking for sys/scsicmd.h... no checking sys/scsiio.h usability... no checking sys/scsiio.h presence... no checking for sys/scsiio.h... no -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org
[sane-devel] New magicolor backend for inclusion in git
I can commit on patches 0004 and 0005. * 0004 looks OK to submit at any time. * 0005 - I just updated the files in this area this week based on Fedora 14's version of autofoo tools. Compared to that update, you have a newer version of autoconf then I do but a quite older version of libtool. On top of that, your update overwrote our hand patched ltmain.sh file. If you want to keep it simple, I suggest once you've submitted patch 0003 I can just regenerate all the Makefiles on your behalf. Chris On Thu, Jan 6, 2011 at 11:30 AM, Reinhold Kainhofer reinhold at kainhofer.com wrote: As you know, I have been developing a magicolor backend for KONICA MINOLTA magicolor 1690MF devices (possibly also for other devices, but I don't have access to any other KONICA MINOLTA device that uses the same protocol -- The bizhub devices use a different protocol). In my view, it is now in a state so that it can be included in sane-backends (I'm using it regularly with xsane). The git patches can be found at: http://www.fam.tuwien.ac.at/~reinhold/sane/magicolor_backend_patches/ -) The first two (0001 and 0002) are for the sanei_usb functionality already send yesterday, -) 0003 is the actual backend, -) 0004 fixes some compiler warnings in byteorder.h, and -) 0005 includes only the changes from running autoreconf (i.e. no manual code changes!) The backend uses libsnmp (configure check added!) to optionally auto-detect a LAN-connected magicolor device. I have set the timeout to a very low value (a little more than 1 second!), so all systems without a magicolor scanning in the network are not held up by the SNMP auto-detection of the magicolor devices. On the other hand, if the network is really slow, this might mean that we miss an SNMP response that takes longer than 1 second! What do you think of this backend? Cheers, Reinhold -- -- Reinhold Kainhofer, reinhold at kainhofer.com, http://reinhold.kainhofer.com/ ?* Financial Actuarial Math., Vienna Univ. of Technology, Austria ?* http://www.fam.tuwien.ac.at/, DVR: 0005886 ?* LilyPond, Music typesetting, http://www.lilypond.org -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org
[sane-devel] pros and cons regarding --enable-pthread for Linux
On Fri, Aug 27, 2010 at 5:45 AM, Johannes Meixner jsmeix at suse.de wrote: Hello, On Aug 26 20:41 m. allan noah wrote: Pthread is the default on other platforms because they generally don't let open usb device handles pass across a fork(). Since this works on Linux, and threads were so flaky all those years ago, non-threaded became the default for Linux. Honestly, I think it could be changed, but I don't think pthread is as well tested, since most user's use whatever their distro compiles :) Then a question to the Linux distribution people on this list is whether or not there is perhaps already a Linux distribution which compiles sane-backends with the non-default configure setting --enable-pthread. Or was this perhaps done in the past but because of whatever issues it was changed back to the default? I know Fedora does not enable pthreads... but I do not know if its intentional. I think a lot of distributions assume disabled features are that way for a reason and don't bother with enabling them to much unless issues like this come up. Chris
[sane-devel] Testing on Canon MF6580PL
On Wed, Jun 23, 2010 at 4:15 AM, twain43 at hotmail.com wrote: Whooopssorry: my fault. Here they are: redone today. Twain28 Subject: Re: [sane-devel] Testing on Canon MF6580PL From: nicolas0martin at gmail.com To: twain43 at hotmail.com CC: chris at cnpbagwell.com; sane-devel at lists.alioth.debian.org Date: Tue, 22 Jun 2010 19:22:23 +0200 I think you omitted to send the zipped logs with your previous message ? Nicolas Le mardi 22 juin 2010 ? 11:14 +, twain43 at hotmail.com a ?crit : Took longer than usual, but I did as you recommended. 1. Built the libusb version linked and installed it; 2. Exported only PIXMA and scanimaged. Log in the attached zip file; 3. Exported LIBUSB and SANEI USB debug. Log in the attached zip file; No news, I'm a fraid. Twain28 I've been fading in and out of this thread so I do not have total history of your logs. Loooking at only this log, it seems to show scanner working correctly while in test mode. Can you do same debug logs but while doing an actual scan? Does that work now with the patched libusb from twain-sane (does it generate a good pnm file)? Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20100628/e5be85f9/attachment.htm
[sane-devel] Testing on Canon MF6580PL
Oh, OK. So one part of the problem is that pixma debug messages aren't showing up. Can the tester verify as a separate task if scanning is working with twain-sane version of libusb though? It seems it may be working now? On debug output part, I see pixma is using some custom stuff to disable when NDEBUG is present. Seems pretty easy that externally included header defines NDEBUG that pixma is including but not sanei_debug.h. A quick test of that theory is to modify the file backends/pixma_common.h around line 57 and change #ifdef NDEBUG to #ifdef NDEBUGBL. Also, similar change needed at backends/pixma.c line 70. Can On Mon, Jun 28, 2010 at 11:46 AM, Nicolas Martin nicolas0martin at gmail.comwrote: Point is, for whatever reason, Pixma debug logs do not show up even when setting SANE_DEBUG_PIXMA to a correct level, only SANE_DEBUG and SANE_DEBUG_SANEI_USB produce logs, as showed up in the logs twain43 is providing. Just trying to figure out why this happens when compiled in this environment. Nicolas 2010/6/28 Chris Bagwell chris at cnpbagwell.com I've been fading in and out of this thread so I do not have total history of your logs. Loooking at only this log, it seems to show scanner working correctly while in test mode. Can you do same debug logs but while doing an actual scan? Does that work now with the patched libusb from twain-sane (does it generate a good pnm file)? Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20100628/faf328b4/attachment.htm
[sane-devel] Testing on Canon MF6580PL
Here is a suggestion. Since twain-sane binary seems to work better for you, perhaps it has a patched version of libusb that works around some OSX bug. Try comping and installing its version of libusb such that your git sane-backends links to it: http://www.ellert.se/PKGS/libusb-2009-09-10/src/libusb-2009-09-10-svn-macosx.tar.gz If you compile and install with --prefix=/usr/local then be sure and add LDFLAGS=-L/usr/local/lib CPPFLAGS=-I/usr/local/include to your sane configure command line. After that, if its still failing then turn on USB debug options. export LIBUSB_DEBUG 255 export SANE_DEBUG_SANEI_USB 255 Chris On Fri, Jun 18, 2010 at 8:14 AM, twain43 at hotmail.com wrote: I'm a bit lost too If you find somethign else we have not yet tried, just let me know, and I'll see what I can do: for the moment, the only thing I can think of is trying to clean every bit of SANE on my sys and try recompiling and installing with the latest development snapshotsbut I guess they're of no use, if the Pixma module has not changed Twain28 Subject: RE: [sane-devel] Testing on Canon MF6580PL From: nicolas0martin at gmail.com To: twain43 at hotmail.com CC: sane-devel at lists.alioth.debian.org Date: Thu, 17 Jun 2010 22:16:45 +0200 Running a bit out of ideas why it behaves like this and why logs do not come out from the backend. Probably something runs differently on MacOS, but I cannot figure out what exactly. Nicolas Le jeudi 17 juin 2010 ? 16:47 +, twain43 at hotmail.com a ?crit : Already tried. 100% same log (see cut-paste below). :( -- nome.cognome @... Verifica la disponibilit? sui NUOVI dominihttps://signup.live.com/signup.aspx?mkt=it-itrollrs=12lic=1 -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20100618/4faeb569/attachment.htm
[sane-devel] Testing on Canon MF6580PL
Probably binary blob from twain-sane links to a static version of libusb (or distributes its own dyn library) and is compiled different then when self-compiling from git. I believe we've seen several reports of sane on OSX having some USB issues when self compiled. I'm not sure if twain-sane web site gives compile instructions for libusb and sane that its using? Chris On Wed, Jun 16, 2010 at 9:49 AM, Nicolas Martin nicolas0martin at gmail.comwrote: Another strangeness here. I can read this from this log: libusb/darwin.c usb_bulk_transfer: USB pipe is an interrupt pipe. Timeouts will not be used. meaning that Darwin libusb still does not handle interrupt timeouts. So I don't yet understand how the previous scanimage -T could work ! 2010/6/16 Nicolas Martin nicolas0martin at gmail.com Failing at the very beginning of the scan here. At this point, do you have the scanimage -T command running correctly ? i.e., with the PASS statements ? -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20100616/b98e6fa1/attachment.htm
[sane-devel] Canoscan N1240U and Mac OS X 10.6
On Thu, May 27, 2010 at 9:31 AM, Gernot Hassenpflug aikishugyo at gmail.comwrote: On Thu, May 27, 2010 at 9:05 AM, Darren Goossens goossens at rsc.anu.edu.au wrote: Hi Has anyone got this combination to work? Canoscan N1240U and Mac OS X 10.6. Can you tell me how? I have tried the software at www.ellert.se/twain-sane and it gives me sane-find-scanner and that works: found USB scanner (vendor=0x04a9 [Canon], product=0x220e [CanoScan], chip=LM983x?) at libusb:001:003-04a9-220e-ff-00 # Your USB scanner was (probably) detected. It may or may not be supported by # SANE. Try scanimage -L and read the backend's manpage. Now, this is a supported scanner (plustek), but scanimage -L gives me nothing. Says it cannot find a scanner. I think scanner access now impossible! I have tried installing xsane from MacPorts (completely broken as fas as I can tell) any advice welcome Good news is the scanner is perfectly supported (I have one) but the plustek backend, even in older versions of sane (like 1.0.19). Probably you have an issue with library paths, but someone familiar with MacOSX and macports would need to give you advice on SANE with a mac, sorry. Gernot I'm not real sure how USB devices work under OSX but is probably the source of issue. One of the help pages on above twain-sane site (which I recommend reading) says that it installs its packages under /usr/local tree. So you probably want to edit the file /usr/local/etc/sane.d/plustek.conf file. Look at the comments about giving a USB device vendor/product ID's and using auto for device name instead of /dev/usbcanner. Macports should have same file if you prefer it but under /opt instead of /usr/local. Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20100527/97d74056/attachment.htm
[sane-devel] Heads up: various issues with epson2 in 1.0.21
I've been doing some limited testing using Artisan 800 and epson2 over network connection only. Mostly, just loading up various GUI's and doing previews. 1.0.21 snuck into Fedora 13 at last minute so thats how I was trying it out with things like simple-scan and xsane. I looked at Fedora's bugzilla and only 2 new sane+epson reports. One is related to parsing options so could be related. Not sure why I'm not seeing it. Chris On Thu, May 20, 2010 at 11:17 AM, Julien BLACHE jb at jblache.org wrote: Hi, I've started getting reports of regressions in epson2 in 1.0.21, affecting a number of scanners. Has anyone else been getting similar reports? So far I've had reports about the Stylus 3200 and the Perfection 640 being totally broken and a divide-by-zero bug in the options handling code. It's all I've got so far but I'm afraid there's more to come... two reports in a couple days is meaningful in my experience maintaining SANE :/ JB. -- Julien BLACHE http://www.jblache.org jb at jblache.org GPG KeyID 0xF5D65169 -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20100520/2e1407a2/attachment.htm
[sane-devel] scanimage works, xsane does not
Just a guess since you didn't mention working around following. Macports is installed under /opt's and only looks for its libraries under /opts. I think those twain-sane packages are self-contained packages. So probably when you compiled xsane from Macports it pulled in the older sane-backends-1.0.20 from Macports as well. Can't offer the exact fix for getting Macports to use external libraries but recommend pinging the Macport maintainer of sane-backends to update to 1.0.21. That would be simplest. Chris On Mon, May 17, 2010 at 1:22 AM, Darren Goossens goossens at rsc.anu.edu.auwrote: Hi, Sorry if I am not very knowledgeable; I installed scanimage and sane on a Mac running 10.6 and it works from command line. (http://www.ellert.se/twain-sane/) But xsane, which I installed from MacPorts, which took many hours of compiling things, does not find any devices, even though I can scan and view images using the command line. ImageCapture on the mac works as well (ie it finds the scanner). So I am pretty sure the problem must be with xsane. Here is some output edr:~ goossens$ sane-find-scanner # sane-find-scanner will now attempt to detect your scanner. If the # result is different from what you expected, first make sure your # scanner is powered up and properly connected to your computer. # No SCSI scanners found. If you expected something different, make sure that # you have loaded a kernel SCSI driver for your SCSI adapter. found USB scanner (vendor=0x04a9 [Canon], product=0x220e [CanoScan], chip=LM9832/3) at libusb:003:004-04a9-220e-ff-00 # Your USB scanner was (probably) detected. It may or may not be supported by # SANE. Try scanimage -L and read the backend's manpage. # Not checking for parallel port scanners. # Most Scanners connected to the parallel port or other proprietary ports # can't be detected by this program. # You may want to run this program as root to find all devices. Once you # found the scanner devices, be sure to adjust access permissions as # necessary. edr:~ goossens$ scanimage -L device `plustek:libusb:003:004-04a9-220e-ff-00' is a Canon CanoScan N1240U/LiDE30 flatbed scanner -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20100517/1348f090/attachment-0001.htm
[sane-devel] Epson Stylus SX410 is not recognized although it is listed as supported
I think you nailed it. I looked at the link for the PPA and its a snapshot of sane-backends from February and the git commit Olaf found is dated April. So, kltrg, you'll need to get a newer version of sane-backends. Hopefully, you can get the officially released 1.0.21 by now. Chris On Sun, May 16, 2010 at 7:19 PM, Olaf Meeuwissen olaf.meeuwissen at avasys.jpwrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 m. allan noah : That seems more complete, but now we have to wait for someone with more experience with epson protocol I guess I qualify ;-) As Chris mentioned, the call to esci_set_color_correction_coefficients is definitely not right. The odd thing is that it gets called at all. Looking at the code for 1.0.21 (and git master), there is no way that this command should get called. Digging a bit further, this has been fixed in http://git.debian.org/?p=sane/sane-backends.git;a=commitdiff;h=e75d6e5e0cb223beb47b6756c56c4798082fe9d1 which is in 1.0.21. Looks like the other person is using a snapshot from before that. Hope this helps, - -- Olaf Meeuwissen, LPIC-2 FLOSS Engineer -- AVASYS CORPORATION FSF Associate Member #1962 Help support software freedom http://www.fsf.org/jf?referrer=1962 -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iEYEARECAAYFAkvwi5EACgkQt5qrxaZLMnJhiwCgtwBsX6t2ndHf28Yda5pPU2Z+ EYoAnjwgnmNBNb36r7zNyCAu6xNndWIZ =adin -END PGP SIGNATURE- -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20100516/e9a03245/attachment.htm
[sane-devel] Epson Stylus SX410 is not recognized although it is listed as supported
I don't know the answer but probably the failure is related to this part of the log file: [epson2] using built in CCT profile [epson2] esci_set_color_correction_coefficients [epson2] esci_set_color_correction_coefficients: not supported What version of sane-backends are you running? Looking at current git, that code path is protected by an if(hw_supports). Not sure how your getting it unless its old software. Chris On Fri, May 14, 2010 at 12:37 PM, m. allan noah kitno455 at gmail.com wrote: That seems more complete, but now we have to wait for someone with more experience with epson protocol allan On Fri, May 14, 2010 at 1:35 PM, kltrg m.kltrg at googlemail.com wrote: I started from scratch again, started xsane via the commandline redirecting the output to that file, hit preview scan, git the error message, closes that notification window and closed xsane. The logfile is attached. I hope it worked that time. kltrg Am 14.05.2010 18:52, schrieb m. allan noah: I dont see any problems in this file. Are you sure it was not truncated? allan On Fri, May 14, 2010 at 12:46 PM, kltrg m.kltrg at googlemail.com wrote: I attached the logfile to the mail. Thanks for all you efforts! kltrg Am 14.05.2010 18:01, schrieb m. allan noah: well, that looks good, do the same when trying to scan: SANE_DEBUG_EPSON2=255 xsane 2xsane.log go until it gives you an error, and send the log here (compress it if large) allan On Fri, May 14, 2010 at 11:41 AM, kltrg m.kltrg at googlemail.com wrote: I get this: http://pastebin.com/uG81XyKn Am 14.05.2010 14:40, schrieb m. allan noah: Try it with the DEBUG flag I gave before... allan On Fri, May 14, 2010 at 8:07 AM, kltrg m.kltrg at googlemail.com wrote: I added the line and now I get: :~$ scanimage -L device `epson2:libusb:002:008' is a Epson PID 0851 flatbed scanner This sounds good and xsane and gscan2pdf recognize the scanner but I still can't scan. I always get an error message: 'Failed to start scanner: Operation not supported'. -- The truth is an offense, but not a sin -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org
[sane-devel] SANE source compilation problems
On Mon, Apr 26, 2010 at 5:28 AM, Julien BLACHE jb at jblache.org wrote: Chris Bagwell chris at cnpbagwell.com wrote: Hi, Others have replied about the remaining issues once @RESMGR@ typo is Speaking of resmgr, I think we can safely remove any code using it. I doubt it's still in use in SuSE... I'd be very surprised if it was. If no one else objects, I'll happily remove it from Makefiles and source code. That is unless someone else prefers doing it. Chris
[sane-devel] SANE source compilation problems
http://gitorious.org/opensuse/resmgr says resmgr was removed in 11.0. distrowatch.com says openSUSE 11.0 was released 2008-06-19. We are couple months shy of 2 years. By time next sane release it will be well over 2 years. :) Opinions? I'll put it on my slw track for now. Chris On Mon, Apr 26, 2010 at 3:04 PM, m. allan noah kitno455 at gmail.com wrote: When was the last version of suse that used it replaced? If more than two years, kill it. allan On Mon, Apr 26, 2010 at 3:52 PM, Chris Bagwell chris at cnpbagwell.com wrote: On Mon, Apr 26, 2010 at 5:28 AM, Julien BLACHE jb at jblache.org wrote: Chris Bagwell chris at cnpbagwell.com wrote: Hi, Others have replied about the remaining issues once @RESMGR@ typo is Speaking of resmgr, I think we can safely remove any code using it. I doubt it's still in use in SuSE... I'd be very surprised if it was. If no one else objects, I'll happily remove it from Makefiles and source code. ?That is unless someone else prefers doing it. Chris -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org -- The truth is an offense, but not a sin
[sane-devel] SANE source compilation problems
On Sun, Apr 25, 2010 at 3:25 PM, Ilm?rs Poik?ns ilmars at delibero.lv wrote: Hi, I downloaded source from git repositories (sane-backends and sane-frontends) and tired to compile them. I have Ubuntu 9.10, needed dev library packages were there. sane-backends configure, make, make install was fine. sane-frontends configure failed. I traced problem down to /user/local/bin/sane-config having line: LIBS= -ldl -lv4l1?? -lm -ltiff -ljpeg? -lgphoto2 -lgphoto2_port -lexif -lm -lusb?? -lusb? @REGMGR_LIBS@ When I deleted @REGMGR_LIBS@ from that line, then sane-frontends configure run successfully. I didn't care about graphical frontends, so configure disabled GUI. Previously gcc tried to find that file and failed to compile some tests. This part was my fault. I recently fixed oen sane-config bug and introduced another. I've just submitted a fix so if you re-pulled sane-backends from git and reinstall, this part of the problem should go away. Others have replied about the remaining issues once @RESMGR@ typo is fixed. sane-frontend in current git contains some next generation API's that sane-backends has disabled. The released sane-frontends package is the one to use with released and git versions of sane-backends. Chris
[sane-devel] sane-backends release 1.0.21
Thanks very much. Can I assume this patch will be included in your future releases of epkowa backend? Allan, are *.desc updates acceptable at this point of freeze? Thanks, Chris On Mon, Apr 19, 2010 at 6:30 PM, Olaf Meeuwissen olaf.meeuwissen at avasys.jp wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 m. allan noah wrote: [...] Please check your .desc files, and make sure they are up to date with all the scanners your backends support, and make sure none of your supported scanners are listed in unsupported.desc. Here's a diff for epkowa.desc based on the thread[1] started by Chris Bagwell regarding some SCSI models that needed some extra info. ?[1] http://lists.alioth.debian.org/pipermail/sane-devel/2010-March/026209.html Hope this helps, - -- Olaf Meeuwissen, LPIC-2 ? ? ? ? ? FLOSS Engineer -- AVASYS CORPORATION FSF Associate Member #1962 ? ? ? ? ? ? ? Help support software freedom ? ? ? ? ? ? ? ? http://www.fsf.org/jf?referrer=1962 -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.9 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iEYEARECAAYFAkvM56MACgkQt5qrxaZLMnLqQQCeJtdQMEErdmbJUtoIXRdVlwJF WpIAoIIdASCnZNKMLDxi5nNNpIpAnubh =x44g -END PGP SIGNATURE- -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org
[sane-devel] sane-backends release 1.0.21
On Mon, Apr 19, 2010 at 6:30 PM, Olaf Meeuwissen olaf.meeuwissen at avasys.jp wrote: Here's a diff for epkowa.desc based on the thread[1] started by Chris Bagwell regarding some SCSI models that needed some extra info. ?[1] http://lists.alioth.debian.org/pipermail/sane-devel/2010-March/026209.html Thanks for the information! The diff has been commited. Chris
[sane-devel] Description file, Epson, and wildcards
On Mon, Mar 1, 2010 at 11:36 PM, Olaf Meeuwissen olaf.meeuwissen at avasys.jp wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Chris Bagwell wrote: On Fedora, it seems its a common problem that hal or udev gets the permissions for scsi Epson scanners wrong and users can only scan as root. ?They generically fix permissions for scsi scanner class but not processor. ?Their work around is to edit entry for specific scanner in the epson2.desc file and a line as follows each time a user reports a problem: :scsi ESPON Exact Model Name Reported processor The 2010-03-01 git snapshot has ?$ ?grep :scsi doc/descriptions*/ep*.desc ?doc/descriptions/epson.desc::scsi EPSON SCANNER GT-7000 processor ?doc/descriptions/epson.desc::scsi EPSON Perfection1200 processor ?doc/descriptions/epson.desc::scsi EPSON Perfection636 processor ?doc/descriptions/epson.desc::scsi EPSON Perfection1200 processor already. ?Are you talking about other models? Yes, they have a patch for Expression800 in addition. My assumption is it would be an issue with any SCSI-only EPSON scanner? Or is there a small subset of names all models use? Mostly, the epson2.desc values are driven by epkowa developers from their epkowa.desc. ?Would it be possible for epkowa developers to provide a :scsi line for all SCSI devices? Yes. ?No promises on a release time line but we could send a diff to the list and ask someone with the necessary privileges to commit. My understanding is that epson2.desc updates are automated based on epkowa.desc updates. So if you could send this updated file to list when its ready, I can submit it. That would be great! Its brute force fix but it will work nicely. A third option is to provide some sort of hard coded/hand written wild card rules that is added to libsane.fdi and ?libsane.rules for these special cases. ?This is probably the easiest. ?It looks like at east Epson and HP could benefit from this. Do people think this is worth addressing? Anything that makes out of the box scanner device detection and setup so you can scan without root privileges is worth addressing. I'll keep the third option in the back pocket... depending on what time line becomes for option one we can pull it out. Chris
[sane-devel] Description file, Epson, and wildcards
On Tue, Mar 2, 2010 at 5:42 PM, Olaf Meeuwissen olaf.meeuwissen at avasys.jp wrote: Yes, they have a patch for Expression800 in addition. ?My assumption is it would be an issue with any SCSI-only EPSON scanner? ?Or is there a small subset of names all models use? Hadn't heard about that one, I'll add that to epkowa.desc as well then. ?Assume it is for the Expression 800 only and not for its Japanese cousin the GT-9600. Sorry for not including that. The report was for Expression 800 entry. The following is bug report if interested. https://bugzilla.redhat.com/show_bug.cgi?id=456656 To be honest, I don't know what the various EPSON scanners with a SCSI interface report. ?Apparently, most advertise themselves as a scanner device so no special handling should be required. ?Some models however think that they are processors and need the :scsi entry. ?What names these models report, I don't know, so I will have to rely on whatever people report. Mostly, the epson2.desc values are driven by epkowa developers from their epkowa.desc. ?Would it be possible for epkowa developers to provide a :scsi line for all SCSI devices? Yes. ?No promises on a release time line but we could send a diff to the list and ask someone with the necessary privileges to commit. My understanding is that epson2.desc updates are automated based on epkowa.desc updates. As m. allen noah also pointed out, that's living life dangerously. If epkowa.desc can include those :scsi that are known, it would be great. Since the epkowa.desc is used in sane-backend build process, fedora would get your :scsi fixes when users install sane-backends and these reports would reduce some... maybe helping isane out in the process. The epson2.desc entries may need further discussion but it sounds like my initial wish of removing a downstream patch can be addressed sometime in near future via a epkowa.desc update. Chris
[sane-devel] Description file, Epson, and wildcards
Hi all, I'm reviewing Fedora patch set and notice an issue that I'm not sure how to handle within Sane. Has this been discussed in past? It looks like all 3 Epson backends can detect SCSI scanners based on scsi type being processor, vendor ID being EPSON, and the model/product name having 1 of 5 prefixes. This is good to know for next step. On Fedora, it seems its a common problem that hal or udev gets the permissions for scsi Epson scanners wrong and users can only scan as root. They generically fix permissions for scsi scanner class but not processor. Their work around is to edit entry for specific scanner in the epson2.desc file and a line as follows each time a user reports a problem: :scsi ESPON Exact Model Name Reported processor Once thats done, sane's tools will generate a libsane.fdi/libsane.rules that will set permissions correct. Sane is not providing enough details right now it seems. First option to help end users is to start providing that much details. Mostly, the epson2.desc values are driven by epkowa developers from their epkowa.desc. Would it be possible for epkowa developers to provide a :scsi line for all SCSI devices? Second option is to provide some kind of wild card support. For example: :scsi EPSON Perfection* processor This would be pretty difficult to convert into hal and udev yntax though. Its also will cause some confusion when a Perfection 636S rule matches a Perfection 636. A third option is to provide some sort of hard coded/hand written wild card rules that is added to libsane.fdi and libsane.rules for these special cases. This is probably the easiest. It looks like at east Epson and HP could benefit from this. Do people think this is worth addressing? Chris
[sane-devel] scanimage 1.0.20 hangs on Solaris 10/SPARC (pixma backend)
The easy thing I can suggest is to try compiling with --enable-pthread. I recall a past discussion related to libusb and that sometimes causing it to work better. If that doesn't fix things for you then your going to have to debug this yourself. I do not think there are any active Solaris developers on the mailing list. See man san, man sane-usb, and man sane-backend (replace backend with real name) for information on what environment variables you can see to enable debug output. Chris On Sun, Feb 28, 2010 at 2:13 AM, Patrick Simmons linuxrocks123 at netscape.net wrote: Hello all, I'm trying to get SANE 1.0.20 working with a Canon MP210 USB scanner on Solaris/SPARC. ?I'm using the latest Solaris 10 build. ?However, I can't get it to scan. ?When I do scanimage -L, my scanner is shown correctly after a moderately long delay, but scanimage just hangs forever when I try to use it to scan. ?scanimage will also hang forever when run with the --help option, after printing out the help text. ?It's as if it's trying to print the scanner name but ran into trouble. I've followed README.Solaris to the letter, except that I also changed an include path in the Makefile from -I/usr/local/include to -I/usr/include because the md5.h in /usr/local/include won't compile on my machine for some reason. ?I experience the same behavior with the latest GIT snapshot. ?The same behavior happens whether I am root or a user. ?I have verified that the scanimage works with SANE 1.0.20 on Linux/AMD64; in addition, I until recently ran Linux/SPARC on the Solaris 10 machine. ?scanimage worked there with a development snapshot of SANE calling itself SANE 1.1.0. Does anyone know what could be wrong? Thanks, --Patrick
[sane-devel] why hasn't there been a release in a long time ?
I was going to wait a little closer to 1 year anniversary of last release before bringing it up but thats starting to creep up on us. I was also hoping we can get one out in time for Fedora 13 but their freeze is coming up pretty quick. I'll offer to help out as much as possible with the release. I can do my set of platform checks (cygwin, osx, solaris, pulling any Fedora patches that make sense, etc). Basically, I can make sure the steps in release.txt are pain free up to the Announcing Release step. I do not recall seeing any 1.1 API stuff being re-enabled although there was talk of it. Is there any work to be done in that area? Chris On Mon, Feb 8, 2010 at 8:16 AM, m. allan noah kitno455 at gmail.com wrote: I personally will take the great measure of responsibility for this problem. My time has become very limited since I took over the job of making releases. I have also allowed the confusion around a new API prevent us from moving forward with maintenance releases, even though that is an unrelated issue. If someone else wants to step up and schedule/make the release, I'd be happy for the help. Otherwise, let me look at a calendar, and come up with a plan. Thanks for the kick in the pants :) allan On Mon, Feb 8, 2010 at 6:39 AM, Julien BLACHE jb at jblache.org wrote: db db.pub.mail at gmail.com wrote: Hi, I know this is not the fault of sane, however this is causing a large distance between what sane can do and what sane 'does' as shipped. If there's no release, there's nothing new to package. As simple as that. So, yes, it is SANE's fault. I'm not packaging git snapshots because I cannot support them in a stable Debian release. I used to package CVS snapshots when I knew a release was coming up and would be out suited for Debian stable at a time where I could still push it into stable. With the 1.1 API stuff and no release schedule, I can't do that anymore. JB. -- Julien BLACHE ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? http://www.jblache.org jb at jblache.org ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?GPG KeyID 0xF5D65169 -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org -- The truth is an offense, but not a sin -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password ? ? ? ? ? ? to sane-devel-request at lists.alioth.debian.org
[sane-devel] Has anyone managed to get Snapscan backend working with FreeBSD 8.0 (which uses libusb)?
On Tue, Jan 26, 2010 at 3:04 AM, Oliver Schwartz Oliver.Schwartz at gmx.dewrote: Hi all, On 25.01.2010, at 12:26, Torfinn Ingolfsen wrote: In the meantime, another FreeBSD user found a way to fix it, it is just a one-line patch: http://lists.freebsd.org/pipermail/freebsd-usb/2010-January/008172.html With Andrew's patch, my scanner now works: root at kg-v2# sane-find-scanner -q found USB scanner (vendor=0x03f0 [HP], product=0xc202 [Photosmart 8200 series]) at libusb:/dev/usb:/dev/ugen1.2 found USB scanner (vendor=0x06bd [AGFA], product=0x2061 [ Snapscan1212u_2]) at libusb:/dev/usb:/dev/ugen2.2 root at kg-v2# scanimage -L device `snapscan:libusb:/dev/usb:/dev/ugen2.2' is a AGFA SNAPSCAN 1212U_2 flatbed scanner Thanks for the pointer, I'll try to get this fixed in the backend. The code in question was added recently to fix another problem with calculating a unique identifier. Maybe one of the guys who's experienced with autoconf can give me a hint on how to do this the best way. Should I just use #ifdef BSD around the code in question or is there some autoconf magic that would allow me to do a more advanced query? I also have to retest this on OS X - the code in question looks very Linux specific. Kind regards, Oliver I don't think there is any autoconf stuff that can help you... or perhaps what you need is already there. A quick look at sanei_usb_init() and I think its the one generate 2 types of strings depending on if your using libusb 1.0 or not. So at least you should be able to reference HAVE_LIBUSB_1.0 but maybe better to update sanei_usb_init() code to always generate a consistent string? Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20100126/36a9b0fe/attachment.htm
[sane-devel] Epson Perfection 610 Issue with 150 dpi scan
On Sun, Jan 3, 2010 at 7:01 AM, Rainer Dorsch rdorsch at web.de wrote: For the long options issue: I rebuild git sane on another machine (also Debian stable aka Debian 5.0) and the long options error is still there. Even --help has problems. Which library is sane using to process the command line options? Does Debian stable maybe ship with a different version than your system? After running configure, can you look at the file include/sane/config.h? Check if HAVE_GETOPT_H and HAVE_GETOPT_LONG are defined. On debian, I would expect them to be defined. In that case, it uses standard libc version of getopt. I do see one suspicious line in lib/getopt.c and lib/getopt1.c. They do a #include config.h instead of #include ../include/sane/config.h and so in certain cases it may be including the wrong config.h file. In that case, it would mistakenly compile a conflicting version of getopt_long() (which can explain the problem your seeing). Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20100103/55a4b3f3/attachment.htm
[sane-devel] long option format not recognized (Was: Epson Perfection 610 Issue with 150 dpi scan)
There was a typo in first part of Rainer's email. Later on it shows its still not working although its safe fix I should probably submit. I've now been able to reproduce the issue with --help not working with the epson2 backend. The problem occurs when we reach driver specific help and so may be specific to epson2. The all_options doesn't seem to contain the --help option and probably anything from basic_options[] array. Since basic_options[] (which includes --help) is copied to all_options[] after filling in backend options, I suspect something returned by epson2 backend is causing a NULL to be placed in middle and causing getopt_long() to stop processing early. I couldn't find anything obvious like missing options inits from browsing the code. I guess it will need to be brute forced debugged to find where the NULL is coming from. Chris On Sun, Jan 3, 2010 at 5:06 PM, m. allan noah kitno455 at gmail.com wrote: good call chris- you want to submit a fix? allan On Sun, Jan 3, 2010 at 4:14 PM, Rainer Dorsch rdorsch at web.de wrote: Hi Chris, Am Sonntag, 3. Januar 2010 schrieb Chris Bagwell: On Sun, Jan 3, 2010 at 7:01 AM, Rainer Dorsch rdorsch at web.de wrote: For the long options issue: I rebuild git sane on another machine (also Debian stable aka Debian 5.0) and the long options error is still there. Even --help has problems. Which library is sane using to process the command line options? Does Debian stable maybe ship with a different version than your system? After running configure, can you look at the file include/sane/config.h? Check if HAVE_GETOPT_H and HAVE_GETOPT_LONG are defined. On debian, I would expect them to be defined. In that case, it uses standard libc version of getopt. As you expected, it is there: rd at blackbox:~/SW.nobackup/sane-backends$ grep HAVE_GETOPT include/sane/config.h #define HAVE_GETOPT_H 1 #define HAVE_GETOPT_LONG 1 rd at blackbox:~/SW.nobackup/sane-backends$ I do see one suspicious line in lib/getopt.c and lib/getopt1.c. They do a #include config.h instead of #include ../include/sane/config.h and so in certain cases it may be including the wrong config.h file. In that case, it would mistakenly compile a conflicting version of getopt_long() (which can explain the problem your seeing). Hmm...that does solve the problem: I did this modification rd at blackbox:~/SW.nobackup/sane-backends$ git diff diff --git a/lib/getopt.c b/lib/getopt.c index 9dabb82..b8715e3 100644 --- a/lib/getopt.c +++ b/lib/getopt.c @@ -20,7 +20,7 @@ License along with the GNU C Library; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA. */ -# include config.h +# include ../include/sane/config.h #if !defined(HAVE_GETOPT_H) || !defined(HAVE_GETOPT_LONG) ^L /* This tells Alpha OSF/1 not to define a getopt prototype in stdio.h. diff --git a/lib/getopt1.c b/lib/getopt1.c index 5ace00a..082998c 100644 --- a/lib/getopt1.c +++ b/lib/getopt1.c @@ -17,7 +17,7 @@ License along with the GNU C Library; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA. */ -#include config.h +#include ../include/sane/config.h #if !defined(HAVE_GETOPT_H) || !defined(HAVE_GETOPT_LONG) ^L #ifdef HAVE_CONFIG_H rd at blackbox:~/SW.nobackup/sane-backends$ recompiled, set the paths again rd at blackbox:~/SW.nobackup/sane-backends$ export |grep PATH declare -x LD_LIBRARY_PATH=/opt/sane-backends-091230/lib declare -x PATH=/opt/sane-backends-091230/bin:/usr/local/bin:/usr/bin:/bin:/usr/games rd at blackbox:~/SW.nobackup/sane-backends$ and --help is still not recognized rd at blackbox:~/SW.nobackup/sane-backends$ which scanimage /opt/sane-backends-091230/bin/scanimage rd at blackbox:~/SW.nobackup/sane-backends$ and --help is still not recogniced rd at blackbox:~/SW.nobackup/sane-backends$ scanimage --help Usage: scanimage [OPTION]... Start image acquisition on a scanner device and write image data to standard output. Parameters are separated by a blank from single-character options (e.g. -d epson) and by a = from multi-character options (e.g. --device-name=epson). -d, --device-name=DEVICE use a given scanner device (e.g. hp:/dev/scanner) --format=pnm|tiff file format of output file -i, --icc-profile=PROFILE include this ICC profile into TIFF file -L, --list-devices show available scanner devices -f, --formatted-device-list=FORMAT similar to -L, but the FORMAT of the output can be specified: %d (device name), %v (vendor), %m (model), %t (type), %i (index number), and %n (newline) -b, --batch[=FORMAT] working in batch mode, FORMAT is `out%d.pnm
[sane-devel] Fwd: sane_read: Device busy with Canon PIXMA MX310 and sane 1.0.20-4ubuntu3
On Thu, Dec 3, 2009 at 3:34 PM, Nicolas Martin nicolas0martin at gmail.comwrote: Nicolas, I'm curious why you didn't recommend the following instead: make distclean ./configure --enable-pthread Yes Chris, you are right, I did not mention that, this adds the correct -lpthread option to the link statement, and should solve the problem hereunder. However, the export CFLAGS=-DUSE_PTHREAD statement is necessary for me, before running the ./configure ... , otherwise, it will compile with the forked option, as the USE_PTHREAD flag stays unset. So your saying that, for you, running make distclean; ./configure --enable-pthread; make doesn't correctly have USE_PTHREAD defined? If so, do you mind emailing me your config.log and include/src/config.h? I'll see what I can do to get it working. It works for me... a quick code review shows snapscan-usb.c a little suspicious because its not including ../include/sane/config.h but is checking for USE_PTHREAD. I didn't follow its include chain very closely though since we are talking about pixma. Thanks, Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20091203/4d0c5b9d/attachment.htm
[sane-devel] Fwd: sane_read: Device busy with Canon PIXMA MX310 and sane 1.0.20-4ubuntu3
Sorry, meant to reply to mailing list. -- Forwarded message -- From: Chris Bagwell ch...@cnpbagwell.com Date: Wed, Dec 2, 2009 at 7:00 PM Subject: Re: [sane-devel] sane_read: Device busy with Canon PIXMA MX310 and sane 1.0.20-4ubuntu3 To: Rogier Eggers rogiereggers at gmail.com On Wed, Dec 2, 2009 at 5:37 PM, Rogier Eggers rogiereggers at gmail.comwrote: Hi Nicolas, Ok, so indeed, the git version without pthtread works flawlessy (see attached log). However, I so far didn't manage to build sane-backends with pthread. After exactly following your instructions, I get the message below. This is with no ubuntu sane packages installed. Can anybody tell me which dev package I miss to build sane with pthread? Sorry, I can't answer name of debian pthread-devel package but I think you also have some sort of configure error. I don't see the -lpthread during the link. Can you reverify that you did the following line? export LIBS=-lpthread If that doesn't work then try adding this after the make distclean step but before configure step: export PTHREAD_LIBS=-lpthread Nicolas, I'm curious why you didn't recommend the following instead: make distclean ./configure --enable-pthread Does that not work correctly under debian with git? Anything I can do to help get that working? Perhaps it was just some past confusion in debian from having to use --disable-fork to get pthread working and that not being obvious? Chris Regards, Rogier Error message during build process: [snip] rm -fr .libs/libsanei.a .libs/libsanei.la ar cru .libs/libsanei.a .libs/sanei_ab306.o .libs/sanei_constrain_value.o .libs/sanei_init_debug.o .libs/sanei_net.o .libs/sanei_wire.o .libs/sanei_codec_ascii.o .libs/sanei_codec_bin.o .libs/sanei_scsi.o .libs/sanei_config.o .libs/sanei_config2.o .libs/sanei_pio.o .libs/sanei_pa4s2.o .libs/sanei_auth.o .libs/sanei_usb.o .libs/sanei_thread.o .libs/sanei_pv8630.o .libs/sanei_pp.o .libs/sanei_lm983x.o .libs/sanei_access.o .libs/sanei_tcp.o .libs/sanei_udp.o .libs/sanei_magic.o .libs/sanei_jpeg.o ranlib .libs/libsanei.a creating libsanei.la (cd .libs rm -f libsanei.la ln -s ../libsanei.la libsanei.la) make[1]: Leaving directory `/home/rogier/Software/Src/sane-backends-git20091201/sanei' Making all in backend make[1]: Entering directory `/home/rogier/Software/Src/sane-backends-git20091201/backend' make all-am make[2]: Entering directory `/home/rogier/Software/Src/sane-backends-git20091201/backend' /bin/bash ../libtool --silent --tag=CC --mode=compile gcc -DHAVE_CONFIG_H -I. -I../include/sane -I/usr/local/include -I. -I. -I../include -I../include -DLIBDIR=/usr/lib/sane -DBACKEND_NAME=dll -DPATH_SANE_CONFIG_DIR=/etc/sane.d -DPATH_SANE_DATA_DIR=/usr/share -DPATH_SANE_LOCK_DIR=/var/lock/sane -DV_MAJOR=1 -DV_MINOR=0 -DUSE_PTHREAD -W -Wall -Wcast-align -Wcast-qual -Wmissing-declarations -Wmissing-prototypes -Wpointer-arith -Wreturn-type -Wstrict-prototypes -pedantic -ansi -MT libdll_la-dll.lo -MD -MP -MF .deps/libdll_la-dll.Tpo -c -o libdll_la-dll.lo `test -f 'dll.c' || echo './'`dll.c dll.c: In function ?load?: dll.c:535: warning: ISO C forbids conversion of object pointer to function pointer type dll.c:560: warning: ISO C forbids conversion of object pointer to function pointer type mv -f .deps/libdll_la-dll.Tpo .deps/libdll_la-dll.Plo /bin/bash ../libtool --silent --tag=CC --mode=link gcc -DUSE_PTHREAD -W -Wall -Wcast-align -Wcast-qual -Wmissing-declarations -Wmissing-prototypes -Wpointer-arith -Wreturn-type -Wstrict-prototypes -pedantic -ansi -Wl,-z,defs -o libdll.la libdll_la-dll.lo /bin/bash ../libtool --silent --tag=CC --mode=link gcc -DUSE_PTHREAD -W -Wall -Wcast-align -Wcast-qual -Wmissing-declarations -Wmissing-prototypes -Wpointer-arith -Wreturn-type -Wstrict-prototypes -pedantic -ansi -Wl,-z,defs -rpath '/usr/lib' -version-number 1:0:21 -o libsane.la -rpath /usr/lib libsane_la-dll-s.lo ../lib/liblib.la libdll.la sane_strstatus.lo ../sanei/sanei_init_debug.lo ../sanei/sanei_constrain_value.lo ../sanei/sanei_config.lo ../sanei/sanei_config2.lo ../sanei/sanei_usb.lo ../sanei/sanei_scsi.lo ../sanei/sanei_pv8630.lo ../sanei/sanei_pp.lo ../sanei/sanei_thread.lo ../sanei/sanei_lm983x.lo ../sanei/sanei_access.lo ../sanei/sanei_net.lo ../sanei/sanei_wire.lo ../sanei/sanei_codec_bin.lo ../sanei/sanei_pa4s2.lo ../sanei/sanei_ab306.lo ../sanei/sanei_pio.lo ../sanei/sanei_tcp.lo ../sanei/sanei_udp.lo -ldl -lm -lieee1284 -ltiff -ljpeg -lgphoto2 -lgphoto2_port -lexif -lm -lusb -lusb ../sanei/.libs/sanei_thread.o: In function `sanei_thread_kill': sanei_thread.c:(.text+0xe5): undefined reference to `pthread_cancel' ../sanei/.libs/sanei_thread.o: In function `sanei_thread_begin': sanei_thread.c:(.text+0x302): undefined reference to `pthread_create' ../sanei/.libs/sanei_thread.o: In function `sanei_thread_sendsig': sanei_thread.c:(.text+0x3ac): undefined reference to `pthread_kill
[sane-devel] Fwd: Fw: SANE compile problem on Solaris
Meant for this to go to the mailing list for future reference. -- Forwarded message -- From: Chris Bagwell ch...@cnpbagwell.com Date: Fri, Sep 25, 2009 at 10:04 AM Subject: Re: Fw: [sane-devel] SANE compile problem on Solaris To: David Price djprice at mindspring.com On Fri, Sep 25, 2009 at 9:43 AM, David Price djprice at mindspring.com wrote: Chris, Thanks for your reponse and suggestions. I have tried both suggestions, both I still cannot get it compiled but with different errors: First, I downloaded the latest backend version from git, it made it a lot further in the build but then failed with: make[2]: Entering directory `/export/home/price/download/sane/git/install.git20090925/sane-backends-git20090925/backend' /bin/bash ../libtool --silent --tag=CC --mode=link gcc -I/usr/sfw/include -W -Wall -Wcast-align -Wcast-qual -Wmissing-declarations -Wmissing-prototypes -Wpointer-arith -Wreturn-type -Wstrict-prototypes -pedantic -Wl,-z,defs -rpath '/opt/sane/lib/sane' -version-number 1:0:21 -module -L/usr/sfw/lib -o libsane-epjitsu.la libsane_epjitsu_la-epjitsu-s.lo ../lib/liblib.la libepjitsu.la../sanei/sanei_init_debug.lo ../sanei/sanei_constrain_value.lo ../sanei/sanei_config.lo sane_strstatus.lo ../sanei/sanei_usb.lo -lm Undefined first referenced symbol in file roundf ./.libs/libepjitsu.a(libepjitsu_la-epjitsu.o) ld: fatal: Symbol referencing errors. No output written to .libs/libsane-epjitsu.so.1.0.21 collect2: ld returned 1 exit status make[2]: *** [libsane-epjitsu.la] Error 1 make[2]: Leaving directory `/export/home/price/download/sane/git/install.git20090925/sane-backends-git20090925/backend' make[1]: *** [all] Error 2 make[1]: Leaving directory `/export/home/price/download/sane/git/install.git20090925/sane-backends-git20090925/backend' make: *** [all-recursive] Error 1 I ran into this myself as well. What I kinda remember was my LD_LIBRARY_PATH was set to something that was giving priority to libm.so.1; which doesn't defined roundf(); instead of libm.so.2. I think I had partially set up my environment to use an older SUNWspro (Sun's own compiler) and that was causing the issues; even though I was compiling with GCC. Try setting up a shell that doesn't have visibility to /opt/SUNWspro/lib or gives priority to /lib first (using LD_LIBRARY_PATH). Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090925/d2a850bc/attachment-0001.htm
[sane-devel] Installing Problems
On Mon, Sep 21, 2009 at 9:03 AM, carterbueford at gmx.net wrote: Hi, please take a look at the attached logfile. There you?ll find the complete aoutput of the make-command. Thanks Oliver OK, based on your log, it can not find the definitions for inb and outb functions for the qcam backend. What OS are you compiling on? Also, can you send your config.log directly to me? It should be in the directory where you ran ./configure from. There is some logic in acinclude.m4 to disable qcam backend if ioperm() and portaccess() are not defined... but it looks slightly buggy how it was phrased (should have been an || I think). And it doesn't check specifically for inb and outb; which seems reasonable at least under Linux but perhaps not on other platforms. -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090921/df93caf2/attachment.htm
[sane-devel] Installing Problems
On Mon, Sep 21, 2009 at 9:17 AM, Chris Bagwell chris at cnpbagwell.com wrote: There is some logic in acinclude.m4 to disable qcam backend if ioperm() and portaccess() are not defined... but it looks slightly buggy how it was phrased (should have been an || I think). And it doesn't check specifically for inb and outb; which seems reasonable at least under Linux but perhaps not on other platforms. OK, the check in acinclude.m4 is correct (it is an either/or thing). I looks like its probably that qcam uses outb and inb but doesn't link in sanei_pio. That file appears to have some logic to handle platforms that do not define those functions. I'll submit a fix to backends/Makefile.am to start linking this in. Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090921/74004fb5/attachment.htm
[sane-devel] Installing Problems
On 09/20/2009 11:19 AM, carterbueford at gmx.net wrote: Hi, I get the following error during installation of sane-server on my mashine: collect2: ld returned 1 exit status make[2]: *** [libsane-qcam.la] Error 1 make[2]: Leaving directory `/opt/install_dir/sane_install/sane-backends-1.0.20/backend' make[1]: *** [all] Error 2 make[1]: Leaving directory `/opt/install_dir/sane_install/sane-backends-1.0.20/backend' make: *** [all-recursive] Error 1 Any suggestions? Thanks. Oliver Probably some sort of unresolved symbols or libraries not found. What does the text before what you included say? Should mention the library/symbols having issues. Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090920/509df3b1/attachment.htm
[sane-devel] Sane Error
On Fri, Jun 26, 2009 at 10:55 AM, Michael Cronenworth mike at cchtml.comwrote: David Solomon on 06/26/2009 10:45 AM wrote: Is there any way to disable the usage globally for testing? If you search through the mailing list archives you should find my MinGW patches (still don't seem to be accepted by the maintainers?) and compile a native libsane.dll unless you require saned. Michael, I was going to attempt to submit at least non-issue chunks of your patches (for example, anything under lib/ which is to support multiple platforms anyways). Last email I saw from you, you mentioned going to re-update your patches so I held off. Mind resending the latest? I've also now installed mingw cross compiler on Fedora 11 box so can help verify things. Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090626/5005df91/attachment.htm
[sane-devel] Sane Error
On Fri, Jun 26, 2009 at 10:45 AM, David Solomon dsolomon at rx30.com wrote: Is there any way to disable the usage globally for testing? David You can try to change to pthreads instead of fork as an easy test by using a configure option. Not sure if it will make a difference. Are you using 1.0.20 or git? For 1.0.20: configure --disable-fork-process For git: --enable-pthreads Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090626/31017af2/attachment.htm
[sane-devel] epson2 updated again
OK, I'm not exactly sure what is different from my last test but now latest update of git and I can network scan great with my Artisan 800. I just tried a hand full of resolutions and both flatbed and ADF. All looked visually fine. Thanks for your work on this! Now, I guess I should go off and figure out what I did to sanei_tcp_read() in return. Chris On 06/02/2009 06:28 PM, Alessandro Zummo wrote: Updated: - added more devices - updated description file - rescan devices on sane_get_devices Still in the pipeline: - color correction profiles - infrared support - support for warmup condition
[sane-devel] byteorder.h
If no one else gives it a go, I'll have a hand at updating the autoconf macro. I've not much experience with cross platform locations of those defines. I see in my Fedora 11 Linux box that alot of those duplicates are defined in endian.h (although perhaps hidden by some #ifdef's) and the closest the macro looks for them is in machine/endian.h. I don't know any history between those two locations or when a needed __USE_BSD is defined by compilers. Chris On 06/01/2009 05:00 PM, Alessandro Zummo wrote: it seems the autoconf generated byteorder.h overrides functions defined in the usual includes, generating messages like this one: ../include/byteorder.h:32:1: warning: htobe16 redefined In file included from /usr/include/sys/types.h:217, from /usr/include/sys/uio.h:24, from /usr/include/sys/socket.h:28, from /usr/include/netinet/in.h:25, from ../include/sane/sanei_tcp.h:23, from epson2_net.c:20: any suggestion?
[sane-devel] Epson PX700W/Artisan 700 network scanning
On Thu, May 28, 2009 at 2:46 PM, Dominik Wnek dominalien at netscape.netwrote: Alessandro Zummo pisze: On Thu, 28 May 2009 08:43:21 +0900 Alesh Slovak alesh.slovak at avasys.jp wrote: iscan limits itself to the epkowa backend, but Xsane was probably using the epson2 backend, which supports the necessary network protocol, but it seems like it doesn't have support for your particular model. Maybe Alessandro (epson2 backend developer) can help you here. If it has been detected by epson2 there are chances that we can make it work. I'd need a debug log. Hello, Thank you very much for your answer. I'll be more than happy to assist with this. Please tell me what to do. Best regards, Dominik Before running your application (xsane, scanimage, etc), run the following from a bash shell: export SANE_DEBUG_EPSON2=128 export SANE_DEBUG_EPSON2_NET=128 Then run something like following to capture debug logs: scanimage 2 scanning.log Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090528/59b7f9e3/attachment.htm
[sane-devel] Warning during configuration of backends 1.0.20 on Mac OS X
I see the issue now but do not know a way to fix it. BTW, the issue is harmless but the scary looking message is built into the tools we use and not able to be disabled. Anyways, reviewing sane's OS X SCSI code, I see any time they compile something, there is a #undef VERSION that conflicts with the OS's header files. I don't know an easy way to undef before checking for header files. Chris On Tue, May 26, 2009 at 2:36 AM, George Katselis katselis at comcast.netwrote: Hi, Devs. I'm using Mac OS 10.5.7.9J61 on a Mac Book 2.16 C2D. I'm trying to compile the SANE Backends 1.0.20 from debian.org. During configure, this warning was displayed, with the message to report it to you. Hope it's helpful! If I can provide more info, please let me know. configure: WARNING: IOKit/scsi/SCSITaskLib.h: present but cannot be compiled configure: WARNING: IOKit/scsi/SCSITaskLib.h: check for missing prerequisite headers? configure: WARNING: IOKit/scsi/SCSITaskLib.h: see the Autoconf documentation configure: WARNING: IOKit/scsi/SCSITaskLib.h: section Present But Cannot Be Compiled configure: WARNING: IOKit/scsi/SCSITaskLib.h: proceeding with the preprocessor's result configure: WARNING: IOKit/scsi/SCSITaskLib.h: in the future, the compiler will take precedence configure: WARNING: ## - ## configure: WARNING: ## Report this to sane-devel at lists.alioth.debian.org ## configure: WARNING: ## - ## -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090527/bd70ee14/attachment.htm
[sane-devel] [sane-commit] [SCM] SANE backends - scanner drivers branch, master, updated. gitconversion-23-g3ba5515
On 05/11/2009 12:12 AM, stef wrote: Le Sunday 10 May 2009 04:51:56 chris at cnpbagwell.com, vous avez ?crit : The following commit has been merged in the master branch: commit 3ba551592c3ad71f82f6b0b2e19c248cc8fe40c1 Author: chris at cnpbagwell.comchris at cnpbagwell.com Date: Sat May 9 21:50:35 2009 -0500 Adding changelog for recent commit. diff --git a/ChangeLog b/ChangeLog index 0146403..e68983f 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,6 @@ +2009-05-09 Chris Bagwellcbagwell-guest at users.alioth.debian.org +* */*.c: standardize including local sane include files. + 2009-05-09 Nicolas Martinnicols-guest at users.alioth.debian.org * backend/pixma_mp150.c, doc/descriptions/pixma.desc: Fixed Pixma MP600 and MP600R for 2400 dpi scan, updated their descriptions. Hello, I have a couple of problems with commits regarding this entry. First I don't find */*.c a useful changelog entry. Sorry, bit of laziness but hoped it read correctly in that every *.c file was touched in some way (at least by a sed script) and so this commit should be treated suspect as such. Second, I really think that active maintainers should be asked on the sane- devel list before changing the code they responsible for. Then go for the change if there is no answer in a decent amount of time. This is going to come up from time to time... There are housekeeping items I'd like to do such as these header bugfixes/cleanups (depending on file if its a bug), standardizing on stdint.h types when portability becomes an issue, etc. I'd think this particular patch would bore people to death reading it but if authors prefer a heads up then not much issue from my side to publish diff to mailing list first. BTW is there any way to include the diffstat in commit messages ? I think this would make them more readable. I second that... I preferred the diffstat over actual diffs if thats possible. Another would be to have the subject line contain diff summary instead of the current subject line its using. Chris Regards, Stef -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090511/4991afb8/attachment.htm
[sane-devel] [sane-commit] [SCM] SANE backends - scanner drivers branch, master, updated. gitconversion-23-g3ba5515
Yes, I would also appreciate if confirmation or clarification when a rebase can be used. I probably mis-interrepted Julien's useful email to mean a rebase only works when remote server has changes that do not affect your local commits; else a merge is needed. Chris On Mon, May 11, 2009 at 8:02 AM, m. allan noah kitno455 at gmail.com wrote: and here we have a merge commit like julien warned about, where my most recent (large) commit to canon_dr shows up again in this commit. IIRC, we should not use git pull, but instead, get fetch and git rebase to prevent this? not directed at chris particularly, git and i dont seem to get along yet, either allan On Sat, May 9, 2009 at 10:51 PM, chris at cnpbagwell.com chris at cnpbagwell.com wrote: The following commit has been merged in the master branch: commit 07c5723cab9571524db586c3319eb40a93568107 Merge: 4d9e8138af0f399e72599631854fa266cb18a4da 64621621c1e3b0163dbd62d3071df028d2cf0b34 Author: chris at cnpbagwell.com chris at cnpbagwell.com Date: Sat May 9 21:40:55 2009 -0500 Merge branch 'master' of ssh:// cbagwell-guest at git.debian.org/git/sane/sane-backends diff --combined backend/canon_dr.c index 44cd9f3,b5d5ab3..994ee0e --- a/backend/canon_dr.c +++ b/backend/canon_dr.c @@@ -56,8 -56,9 +56,9 @@@ Section 2 - sane_init, _get_devices, _open friends Section 3 - sane_*_option functions Section 4 - sane_start, _get_param, _read friends -Section 5 - sane_close functions -Section 6 - misc functions +Section 5 - calibration functions +Section 6 - sane_close functions +Section 7 - misc functions Changes: v1 2008-10-29, MAN @@@ -182,6 -183,19 +183,19 @@@ v26 2009-04-14, MAN (SANE 1.0.20) - return cmd status for reads on sensors - allow rs to adjust read length for all bad status responses + v27 2009-05-08, MAN + - bug fix in read_panel() + - initialize vars in do_usb_cmd() + - set buffermode off by default + - clear page counter during init and sane_start() + - eject previous page during init and sane_start() + - improved SSM_BUFF macros + - moved set_window() to after ssm-*() + - add coarse calibration (AFE offset/gain per-channel exposure) + - add fine calibration (per-cell offset/gain) + - free image and fine cal buffers in sane_close() + - compare page counter of small scanners only in non-buffered mode + - add back-side gray mirroring code for DR-2580C -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090511/b4edefd4/attachment.htm
[sane-devel] [sane-commit] [SCM] SANE backends - scanner drivers branch, master, updated. gitconversion-23-g3ba5515
On 05/10/2009 01:28 PM, m. allan noah wrote: On Sun, May 10, 2009 at 2:22 PM, Julien BLACHEjb at jblache.org wrote: Ilia Sotnikovhostcc at gmail.com wrote: Hi, Could you please describe in what matter using relative paths to header files (eg. #include ../include/sane/sane.h) is better in comparision with -I../include command line option? Just to have cleaner picture... This way you don't pick up installed SANE include files in the standard include paths. see recent thread about freebsd build issues. allan A brief summary: Using #include sane/*.h and -I for header files that can also be found in /usr/include is not reliable as some cpp's will search -I paths as last option in that case; thus preferring the /usr/include version. So this is a problem with sane/sane.h and sane/saneopts.h but I also applied a chosen solution to all internal header files for consistency. Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090510/f41c1ccf/attachment.htm
[sane-devel] sane-backends 1.0.20 Released
On Mon, May 4, 2009 at 9:03 AM, Julien BLACHE jb at jblache.org wrote: Chris Bagwell chris at cnpbagwell.com wrote: Hi Chris, Are you referring to sane-backends? I can't find a --enable-pthreads option. Also, I tried for the first time the --enable-fork-process option but still had no compile issues. Yes, that's sane-backends and the option is actually --disable-fork-process (which does exactly what --enable-pthreads would do if it existed ;)) It switches sanei_thread from using fork() to using pthreads for reader processes. Ah, thats the part I was missing and can now reproduce. Some of that double negative logic in acinclude.m4 was confusing me and I thought my test on Darwin would have fully tested it. I can have a fix ASAP although I guess I'll need to hold off on submitting until git is ready. BTW, temporary work around for those that need it: ./configure --disable-fork-process LDFLAGS=-lpthread Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090504/d66c8a8a/attachment.htm
[sane-devel] compile sane snapshot under Freebsd 7.1
On Mon, May 4, 2009 at 2:22 PM, Louis Lagendijk louis at lagendijk.xs4all.nlwrote: On Sat, 2009-05-02 at 15:59 -0500, Chris Bagwell wrote: canon_dr.c picks the wrong saneopts.h (from /usr/local/sane). possibly other sources do the same, but I for now simply removed the old sane includes most backends seem to include the system sane files instead of the src ones. perhaps we should change that at some point? I didn't worry about this because I assumed our -I options should always force it to look for internal versions first. AM_CPPFLAGS = -I. -I$(srcdir) -I$(top_builddir)/include -I $(top_srcdir)/include -DLIBDIR=$(libdir)/sane Somehow your still getting -I/usr/local/include with higher priority. How was that specified? ./configure CPPFLAGS=-I/usr/local/include? Can you look at what command line options are being used to compile? You will need to edit backend/Makefile and remove the --silent from $LIBTOOL. Generating umax1220u.conf from umax1220u.conf.in Generating umax.conf from umax.conf.in Generating umax_pp.conf from umax_pp.conf.in Generating v4l.conf from v4l.conf.in Generating xerox_mfp.conf from xerox_mfp.conf.in Generating dll.conf from dll.conf.in Generating saned.conf from saned.conf.in gmake all-am gmake[1]: Entering directory `/usr/ports/graphics/sane-backends/sane-backends/backend' rm -f dll-s.c ln -s ./stubs.c dll-s.c /bin/sh ../libtool --tag=CC --mode=compile gcc -DHAVE_CONFIG_H -I. -I../include/sane -I/usr/local/include -I. -I. -I../include -I../include -DLIBDIR=/usr/local/lib/sane -DBACKEND_NAME=dll -D_REENTRANT -DPATH_SANE_CONFIG_DIR=/usr/local/etc/sane.d -DPATH_SANE_DATA_DIR=/usr/local/share -DPATH_SANE_LOCK_DIR=/usr/local/var/lock/sane -DV_MAJOR=1 -DV_MINOR=1 -g -O2 -W -Wall -Wcast-align -Wcast-qual -Wmissing-declarations -Wmissing-prototypes -Wpointer-arith -Wreturn-type -Wstrict-prototypes -pedantic -ansi -MT libsane_la-dll-s.lo -MD -MP -MF .deps/libsane_la-dll-s.Tpo -c -o libsane_la-dll-s.lo `test -f 'dll-s.c' || echo './'`dll-s.c mkdir .libs The above confirms we are setting up -I options correctly. I believe I see the issue now... We have 3 versions of including saneopts.h in various files. #include sane/getopts.h #include sane/getopts.h #include ../include/sane/getopts.h I had trouble finding an official definition of difference between and but I finally found in GNU's cpp manual that means look in -I places and then standard system header places as a fall back. It then says that looks in current directory but falls back to standard system header places. It makes no mention of referencing -I values. So that seems the issue. We need to get rid of all #include sane/getopts.h and also better to align with just 1 way to do it such as also getting rid of all #include sane/getopts.h as well. I can align this soon if someone else doesn't want to take that up themselves. -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090504/39b78925/attachment-0001.htm
[sane-devel] Git repositories are up
On Mon, May 4, 2009 at 12:02 PM, Julien BLACHE jb at jblache.org wrote: Hi, The git repositories for SANE are up and running on alioth. Just had to make a submission to have an excuse to try git for the first time. So far I'm really liking all those commands you can do without having to contact the server. Perhaps an additional set of tips for us first timers trying to simulate cvs operations. It may save you a few minutes of searching. - git pull is like cvs update. - The two step operation of git commit -a followed by git push is similar to cvs commit. - Seems the standard protocol for commit messages is: [ one line summary of change ] [ empty line ] [ multi-line description of change if required. ] And a few possible issue that I'm sure your aware of Julien. - I didn't receive a commit email. Could be mail server not sending to sender. - I went to sane-commit archive on web page to see if it was sent. The archive is shown as empty. - Sane webpage still points to cvs on download page. Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090504/b279d712/attachment.htm
[sane-devel] sane-backends 1.0.20 Released
On Mon, May 4, 2009 at 10:06 AM, m. allan noah kitno455 at gmail.com wrote: On Mon, May 4, 2009 at 11:03 AM, Julien BLACHE jb at jblache.org wrote: Chris Bagwell chris at cnpbagwell.com wrote: Hi, It switches sanei_thread from using fork() to using pthreads for reader processes. Ah, thats the part I was missing and can now reproduce. Some of that double negative logic in acinclude.m4 was confusing me and I thought my test on Darwin would have fully tested it. Yes, it's a bit convoluted, isn't it? :) If you feel like adding a --enable-pthreads instead, feel free ;) yes- i was just going to say- lets drop --disable-fork and add --enable-pthreads instead. OK, thats what I did. Also, I changed things so that if --disable-pthread is specified that we do not needlessly link in -lpthread or set the CPPFLAGS to re-entrant. I would appreciate a quick glance by someone to make sure I didn't get any logic wrong. Here's to hoping my first commit with git went well. Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090504/458e0321/attachment.htm
[sane-devel] compile sane snapshot under Freebsd 7.1
I'm very glad to hear someone was able to test under FreeBSD. Thats the main platform I do not have access to. Really need to set up a qemu image some day. Replies below. On 05/02/2009 11:14 AM, m. allan noah wrote: On Sat, May 2, 2009 at 11:49 AM, Louis Lagendijk louis at lagendijk.xs4all.nl wrote: - libgphoto2 is not found, solved it temporarily by adding ${GPHOTO2_LDFLAGS} to the end of the GPHOTO2_LIBS line. I know this is an ugly hack Maybe Chris can chime in on that one? Agree its an issue. Got a working patch I'll submit. Also, need a related update to sane-config.in. I'll test a little more and submit tomorrow. canon_dr.c picks the wrong saneopts.h (from /usr/local/sane). possibly other sources do the same, but I for now simply removed the old sane includes most backends seem to include the system sane files instead of the src ones. perhaps we should change that at some point? I didn't worry about this because I assumed our -I options should always force it to look for internal versions first. AM_CPPFLAGS = -I. -I$(srcdir) -I$(top_builddir)/include -I$(top_srcdir)/include -DLIBDIR=$(libdir)/sane Somehow your still getting -I/usr/local/include with higher priority. How was that specified? ./configure CPPFLAGS=-I/usr/local/include? Can you look at what command line options are being used to compile? You will need to edit backend/Makefile and remove the --silent from $LIBTOOL. tools/sane-find-scanner: -lcam missing (added a ${SCSI_LIB} to the LD line ok. Chris? :) Agree its missing. I'll submit patch by tomorrow. Chris Please let me know if more information is needed Louis p.s. the pixma stuff works in first very quick test allan -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090502/8a99125a/attachment.htm
[sane-devel] sane-backends now in feature freeze
On Sun, Apr 19, 2009 at 2:49 PM, m. allan noah kitno455 at gmail.com wrote: It would also be helpful if folks could attempt to build the current sources on any uncommon platforms they have access to, so we can shake out any build system or compiler issues. FYI: I compiled latest CVS successfully with Fedora 11 (beta), Solaris 10+gcc, Cygwin, and Mac OSX (no libusb). Testing was limited to cd testsuite; make test.local. Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090419/74800347/attachment.htm
[sane-devel] Error during compile
[scratching my head] Hmmm, how'd that obvious issue not show up on my system? I've just committed to CVS a fix to stop looking at the template files when generating the HTML pages. Can you update and try again? Chris carterbueford at gmx.net wrote: Hi everybody, I get the following error during compiling the cvs-snapshot: Generating manpage scanimage.1... Generating manpage sane-config.1... Generating manpage sane-find-scanner.1... Generating manpage gamma4scanimage.1... Generating manpage sane.7... Generating manpage saned.8... make[1]: *** No rule to make target `descriptions/template.desc.', needed by `sane-backends.html'. Stop. make[1]: Leaving directory `/opt/saned/sane-backends/doc' make: *** [all-recursive] Error 1 Any idea what is meant by that?? Thank Oliver
[sane-devel] Mac OS X: configure script errors
Yes, I've seen similar issues in all autotools projects. I've seen it most on Mac's as well (don't know if its limited to just that platform though)... The configure scripts don't seem to cope well with spaces in directory names. Not much can be done about this short of fixing the autotools. Chris On Tue, Mar 24, 2009 at 10:58 AM, Peter Schoenrank peter at schoenrank.cawrote: I am using Mac OS X 10.5.6 on an Intel iMac. When I do $ mkdir ~/Desktop/scanning\ from\ the\ command\ line $ cd ~/Desktop/scanning\ from\ the\ command\ line $ curl 'http://alioth.debian.org/snapshots.php?group_id=30186' sane- scm-latest.tar.gz $ open sane-scm-latest.tar.gz $ cd sane-scm-2009-03-24/sane-backends $ export BACKENDS=pixma canon $ ./configure --build=x86-apple-osx I get checking for a BSD-compatible install... /usr/bin/install -c checking whether build environment is sane... yes /bin/sh: /Users/cps/Desktop/scanning: No such file or directory configure: WARNING: `missing' script is too old or missing It would seem that some part of configure is not coping with ~/Desktop/scanning\ from\ the\ command\ line having spaces in its name. Peter - Peter Schoenrank mailto: peter at schoenrank.ca phone: 250-655-6753 -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090324/7ced72ed/attachment.htm
[sane-devel] Mac OS X: Report this ...
Based on your config.log from other email, its failing during check with this error message: configure:17246: checking IOKit/scsi/SCSITaskLib.h usability configure:17263: gcc -c -g -O2 -W -Wall -Wcast-align -Wcast-qual -Wmissing-declarations -Wmissing-prototypes -Wpointer-arith -Wreturn-type -Wstrict-prototypes -pedantic -ansi -D_REENTRANT conftest.c 5 In file included from /System/Library/Frameworks/IOKit.framework/Headers/scsi/SCSITaskLib.h:29, from conftest.c:136: /System/Library/Frameworks/IOKit.framework/Headers/scsi/SCSICmds_INQUIRY_Definitions.h:70: error: syntax error before string constant /System/Library/Frameworks/IOKit.framework/Headers/scsi/SCSICmds_INQUIRY_Definitions.h:80: error: syntax error before '}' token /System/Library/Frameworks/IOKit.framework/Headers/scsi/SCSICmds_INQUIRY_Definitions.h:81: error: syntax error before '*' token /System/Library/Frameworks/IOKit.framework/Headers/scsi/SCSICmds_INQUIRY_Definitions.h:91: error: syntax error before string constant /System/Library/Frameworks/IOKit.framework/Headers/scsi/SCSICmds_INQUIRY_Definitions.h:109: error: syntax error before '}' token In file included from conftest.c:136: /System/Library/Frameworks/IOKit.framework/Headers/scsi/SCSITaskLib.h:594: error: syntax error before 'SCSICmd_INQUIRY_StandardData' configure:17270: $? = 1 So it looks like that header file is not self-supporting. Don't know what else needs to be included first to get rid of failure. I have only occasional access to a Mac and developer docs so can't easily debug. For the most part, this message can be ignored though since its a warning. Chris On Tue, Mar 24, 2009 at 11:07 AM, Peter Schoenrank peter at schoenrank.cawrote: I am using Mac OS X 10.5.6 on an Intel iMac When I run the configure script, I get . . . configure: WARNING: IOKit/scsi/SCSITaskLib.h: present but cannot be compiled configure: WARNING: IOKit/scsi/SCSITaskLib.h: check for missing prerequisite headers? configure: WARNING: IOKit/scsi/SCSITaskLib.h: see the Autoconf documentation configure: WARNING: IOKit/scsi/SCSITaskLib.h: section Present But Cannot Be Compiled configure: WARNING: IOKit/scsi/SCSITaskLib.h: proceeding with the preprocessor's result configure: WARNING: IOKit/scsi/SCSITaskLib.h: in the future, the compiler will take precedence configure: WARNING: ## - ## configure: WARNING: ## Report this to sane-devel at lists.alioth.debian.org ## configure: WARNING: ## - ## . . . Peter - Peter Schoenrank mailto: peter at schoenrank.ca phone: 250-655-6753 -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090324/fd537459/attachment-0001.htm
[sane-devel] Mac OS X: make errors
OK, I have to stick with build issue myself... Sorry, can't help with USB issues. I have only 1 addition suggestion. Try running configure with --prefix=/usr/local/testsane to have it installed in a custom location. I've not fully followed in the other threads if you've ever install a previous copy of sane into the prefix /usr/local... but perhaps libsane is loading some older version of your backend that has these USB issues. Anyways, the --prefix idea can only help I think... but likely is not your issue. Chris On Tue, Mar 24, 2009 at 4:24 PM, Peter Schoenrank peter at schoenrank.cawrote: On 09-Mar-24, at 11:22, Chris Bagwell wrote: This still results in the warnings about IOKit/scsi/SCSITaskLib.h that I previously posted, but as Chris suggested, I ignored those because my scanner is USB not SCSI. $ make check makecheck.log $ make make.log $ sudo make install makeinstall.log $ sane-find-scanner sane-find-scanner.log found USB scanner (vendor=0x04a9 [Canon], product=0x172c [MX850 series]) at libusb:007:002-04a9-172c-00-00 $ scanimage -L scanimage-L.log device `pixma:04A9172C_21B0DA' is a CANON Canon PIXMA MX850 multi-function peripheral $ export SANE_DEBUG_SANEI_USB=255 $ export SANE_DEBUG=255 $ export SANE_DEBUG_PIXMA=20 $ scanimage -d pixma:04A9172C_21B0DA -T scanimage.log And now I am back to where I was on Saturday: scanimage hangs. For those who care, config.log configure.log makecheck.log make.log makeinstall.log sane-find-scanner.log scanimage-L.log scanimage.log are available at http://members.shaw.ca/schoenrank.ca/sane-scm-2009-03-24.2/ Peter - Peter Schoenrank mailto: peter at schoenrank.ca phone: 250-655-6753 -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090324/59b136b4/attachment.htm
[sane-devel] Segmentation Fault
2009/3/3 Russell King rjkfsm at gmail.com Well, I tried that and xsane now complains that it needs sane-backends 1.0.0 or higher. To get here, I had to unmerge sane-backends, sane-frontends and xsane, then to keep portage from bashing them, I masked them. I then downloaded the source for sane-backends and compiled it. That went fine and when I tried the same with xsane, I get: ERROR: SANE-1.0.0 or newer is needed for compiling xsane - if you installed SANE as rpm make sure you also included sane-devel I just did some limited testing with current CVS and xsane- 0.996 and had success. Note: I found an additional typo in sane-config and just submitted a fix to CVS. You'll want to update again. Your error message above is because xsane can not find the sane-config of your CVS sane-backends. Step I did are similar to following: cd sane-backend ./configure --prefix=/usr/local/sane make; make install export PATH=/usr/local/sane/bin:$PATH cd /path/to/src/xsane-0.996 cd src vi xsane-backend-gtk.c comment out all of if(SANE_CAPS_ALWAYS_SETTABLE) at lines 2615-2618 as its not documented by SANE1 API and not in current CVS... It was in sane-1.0.19 for what ever reasons but only used by qcam and v4l backends. I'm sure you'll not notice reduced functionality. cd .. ./configure --prefix=/usr/local/xsane LDFLAGS=-L/usr/local/sane/lib CPPFLAGS=-I/usr/local/sane/include make; make install /usr/local/xsane/bin/xsane -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090303/c081e1d2/attachment.htm
[sane-devel] broken sane-config in sane-backends version 1.1.0
Olaf Meeuwissen wrote: szukw000 at arcor.de writes: I had to make some changes to get a working 'sane-config'. The result: LIBS= -lgphoto2 -lgphoto2_port -lexif -lm -lusb -ldl @GPHOTO2_LDFLAGS@ is empty I do not know whether this is sufficient. sane-backends-1.0.19 has a longer list for LIBS: LIBS=-lusb -lpthread -lm -ljpeg -ltiff -lieee1284 -lgphoto2 -lgphoto2_port -lexif -lm -lusb -ldl OK, thanks for report... I submitted a version of your patches and now I get basically same as 1.0.19: LIBS= -ldl -lv4l1 -lv4l2 -lv4lconvert -lm -ltiff -ljpeg -lgphoto2 -lgphoto2_port -lexif -lm -lusb -lusb TODO: We should really strip out duplicates libraries... TODO2: I don't think that LIBS in most systems should really contain *ANY* values because libsane.so is just libsane-dll.so and doesn't contain any code that requires linking to that stuff. When you dlopen a backend, it also loads its dependencies that were linked into it as needed, right? In that case, sane-config's LIBS should be built based on only whats in PRELOAD_BACKENDS (normally nothing). Chris
[sane-devel] Build system issues.
On Thu, Feb 26, 2009 at 10:07 PM, m. allan noah kitno455 at gmail.com wrote: On Thu, Feb 26, 2009 at 11:00 PM, Chris Bagwell chris at cnpbagwell.com wrote: m. allan noah wrote: On Thu, Feb 26, 2009 at 9:46 PM, Chris Bagwell chris at cnpbagwell.com wrote: Here is an idea to explore... Is there any change your RPM package is replacing ltmain.sh or libtool when it was compiled? I could see how ldconfig might behavior like your seeing if 1.0.19 backends were installed with soname's of the form libsane-backend.so.1 but your 1.1.0 backends were installed with soname's of the form libsane.so.1. Its not 'my' rpm- it is the distro's rpm, which happens to be: Fedora 10, tada! Now you can test it too :) Yep, that appears to be the issue... at least at first glance. This line is in Fedora 10's spec file: # We patched configure.in, so regenerate configure. autoreconf --install --force They should remove that --force option so as not to overwrite our patched ltmain.sh. So using Fedora RPM's and then installing on top with source package will have the unwanted behavior you are seeing. I'll bugzilla that when I have a moment. Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090228/d530d0dd/attachment.htm
[sane-devel] build system issue installing config files
On Sat, Feb 28, 2009 at 3:41 PM, Louis Lagendijk louis at lagendijk.xs4all.nlwrote: hi In order to better keep track of what I installed, I am trying to build an RPM package for Sane on my Fedora 10 box. This fails as the config files do not get installed. This seems to be caused by the install-becfg target in backend/Makefile.am `/home/home1/louis/rpm/BUILD/sane-backends-1.1.0.cvs/backend' rm -f /home/home1/louis/rpm/BUILDROOT/sane-backends-1.1.0.cvs-12.fc10.x86_64/usr/lib64/sane/libsane.* test -z /etc/sane.d || /bin/mkdir -p /etc/sane.d test -z /etc/sane.d/dll.d || /bin/mkdir -p /etc/sane.d/dll.d NOT overwriting abaton.conf in /etc/sane.d... NOT overwriting agfafocus.conf in /etc/sane.d... NOT overwriting apple.conf in /etc/sane.d... I have no idea where to look for the cause. Or am I missing something obvious? In the makefile I see that the install-becfg is triggered by install-dat-hook, which in its turn is called by install-data-am which does $(MAKE) $(AM_MAKEFLAGS) install-data-hook. So is DESTDIR missing from AM_MAKEFLAGS? I am afraid I do not know enough of autoconf to understand all logic involved. Suggestions? Hmmm, I've not had time to work on Fedora's spec file but it will need some changes to work with at least this automake change. The recommended practice is for the spec file to pass in the value of DESTDIR; its not sane-backends that sets it. If Fedora spec would have done that then we'd have been OK. In their spec file look for: %{makeinstall} docdir=%{buildroot}/%{_docdir}/%{name}-%{version} replace that with more recommend practice of: make DESTDIR=%{buildroo} docdir=%{buildroot}/%{_docdir}/%{name}-%{version} install If that gives you problems, then try the following: %{makeinstall} configdir=%{docdir=%{buildroot}/%{_docdir}/%{name}-%{version} configdir=%{buildroot}/etc/sane.d As a side-note: RPM barfs when a package name has more than one dash in it. Would it be possible to change the package version in configure.in so it does not contain a dash, e.g. to 1.1.0cvs? Yep, I run into the same issue. I tend to modify configure.in though and change version # to 1.0.99 (and remove the -cvs) so that if/when Fedora upgrades their RPM's, my custom RPM will not look newer. Not sure if others on list feel its OK to remove the dash from -cvs to help RPM people out in general though? Chris thanks in advance for the help Louis/insta -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090228/2e1511d9/attachment-0001.htm