- Dillo-dev - lists.mailman3.com

Patch: suppress duplicate Fl_Printer::class_id

by james.from.wellington＠gmail.com

Removing the stub on OSX 10.6.8 with fltk 1.3.2 does not make dillo grow. Sizes of dillo image; the first line counts shared libraries, and the second overestimates, perhaps consistently, the size of the to-run code: - with Fl_Printer stub without class_id: $ otool -L src/dillo | wc 14 92 1370 $ ls -l src/dillo -rwxr-xr-x 1 ... staff 2407760 15 Sep 21:21 src/dillo - without Fl_Printer stub: $ otool -L src/dillo | wc 14 92 1370 $ ls -l src/dillo -rwxr-xr-x 1 ... staff 2407760 17 Sep 00:05 src/dillo If it's purpose has been overtaken by events, and it increases the brittleness of the code, then I would be happy for it to be removed. On 9/16/14, eocene <eocene at gmx.com> wrote: > James wrote: >> Building commit 317f010 (two days ago) of dillo, I got a duplicate >> symbol on OSX 10.6.8 with gcc 4.2.1. Here is the diff that suppresses >> it. >> >> ---------------- >> $ hg diff >> diff -r c09473fb7311 dw/fltkplatform.cc >> --- a/dw/fltkplatform.cc Sat Sep 13 09:46:28 2014 +0000 >> +++ b/dw/fltkplatform.cc Sat Sep 13 23:49:26 2014 +1200 >> @@ -37,10 +37,10 @@ >> * any need for currently. This stub prevents that. */ >> class FL_EXPORT Fl_Printer : public Fl_Paged_Device { >> public: >> - static const char *class_id; >> +// static const char *class_id; >> Fl_Printer(void) {}; >> }; >> -const char *Fl_Printer::class_id = "Fl_Printer"; >> +//const char *Fl_Printer::class_id = "Fl_Printer"; > > To refresh my memory of the magnitude of what printer stuff we pull in > without the stub, I commented out the stub -- and saw nothing to do with > Fl_Printer with nm. > > I thought maybe this somehow had to do with the changes that Lauri Kasanen > made to fltk last month like > http://www.fltk.org/newsgroups.php?s10767+gfltk.commit+v10781 > http://www.fltk.org/newsgroups.php?s10787+gfltk.commit+v10787 > but I see that Manolo took out the Fl_Printer::class_id check in > Fl_Text_Display in March 2012. 1.3.2 came out in...let's see... > December 2012. > > So we should probably remove the stub altogether. > > _______________________________________________ > Dillo-dev mailing list > Dillo-dev at dillo.org > http://lists.dillo.org/cgi-bin/mailman/listinfo/dillo-dev >

Sept. 16, 2014

Dillo and Irix

by johannes.hofmann＠gmx.de

On Tue, Sep 16, 2014 at 06:29:27AM +0000, eocene wrote: > > In the typical vsnprintf() (i.e., following the c99 standard), if you > have space for 10 bytes and you want to write 20 bytes (including '\0'), > you return 19, whereas irix returns 9. > > Some options: > > 1. It's possible to make irix and the standard run with a change like > > diff -r 5f63e28334b2 dlib/dlib.c > --- a/dlib/dlib.c Tue Sep 16 00:21:49 2014 +0000 > +++ b/dlib/dlib.c Tue Sep 16 04:59:20 2014 +0000 > @@ -406,11 +406,11 @@ > va_copy(argp2, argp); > n = vsnprintf(ds->str + ds->len, ds->sz - ds->len, format, argp2); > va_end(argp2); > - if (n > -1 && n < ds->sz - ds->len) { > + if (n > -1 && n+1 < ds->sz - ds->len) { > ds->len += n; /* Success! */ > break; > } else if (n > -1) { /* glibc >= 2.1 */ > - n_sz = ds->len + n + 1; > + n_sz = ds->len + n + 2; > } else { /* old glibc */ > n_sz = ds->sz * 2; > } > > ...although it would need tweaking to waste less running time. > > But of course even if dillo is partly about retrocomputing, it'll be > 99% linux+bsd retrocomputing and 1% other-stuff retrocomputing, so... > > 2. Throw an ifdef at it. Dillo has never had many ifdefs, and I'd thought > they were discouraged, but maybe the changes needed for all of that stuff > at the bottom of Compatibility.html like > > SGI O2 running IRIX > Tru64 (OSF1) 4.0 on Alpha > QNX RTP 6.1/x86 (with occasional memory faults out of GTK+) > AIX 4.3 (with some tweaks) > Atari-based 68k-systems running MiNT > Microsoft Windows using Cygwin > RISC OS > > ...just didn't make it back into mainline dillo1? > > 3. We could always borrow fltk's vsnprintf(), which is about 250 lines. > > > I lean toward giving it an ifdef. I would suggest testing for this specific (v)snprintf() behaviour in configure and ifdef based on the result rather than on the OS. There might even be some builtin support for this check in autoconf as the issue is mentioned in [1]. Johannes [1] https://www.gnu.org/software/autoconf/manual/autoconf-2.64/html_node/Functi…

Sept. 16, 2014

Dillo and Irix

by diegel＠work.de

eocene wrote: > 1. It's possible to make irix and the standard run with a change like dilo 3.0.4 works fine with your pach on Irix. Thank you very much for your help! > I lean toward giving it an ifdef. I would also be happy to maintain a patch for Irix dillo. In this case you should make some notes in the README file.

Sept. 16, 2014

Dillo and Irix

by eocene＠gmx.com

In the typical vsnprintf() (i.e., following the c99 standard), if you have space for 10 bytes and you want to write 20 bytes (including '\0'), you return 19, whereas irix returns 9. Some options: 1. It's possible to make irix and the standard run with a change like diff -r 5f63e28334b2 dlib/dlib.c --- a/dlib/dlib.c Tue Sep 16 00:21:49 2014 +0000 +++ b/dlib/dlib.c Tue Sep 16 04:59:20 2014 +0000 @@ -406,11 +406,11 @@ va_copy(argp2, argp); n = vsnprintf(ds->str + ds->len, ds->sz - ds->len, format, argp2); va_end(argp2); - if (n > -1 && n < ds->sz - ds->len) { + if (n > -1 && n+1 < ds->sz - ds->len) { ds->len += n; /* Success! */ break; } else if (n > -1) { /* glibc >= 2.1 */ - n_sz = ds->len + n + 1; + n_sz = ds->len + n + 2; } else { /* old glibc */ n_sz = ds->sz * 2; } ...although it would need tweaking to waste less running time. But of course even if dillo is partly about retrocomputing, it'll be 99% linux+bsd retrocomputing and 1% other-stuff retrocomputing, so... 2. Throw an ifdef at it. Dillo has never had many ifdefs, and I'd thought they were discouraged, but maybe the changes needed for all of that stuff at the bottom of Compatibility.html like SGI O2 running IRIX Tru64 (OSF1) 4.0 on Alpha QNX RTP 6.1/x86 (with occasional memory faults out of GTK+) AIX 4.3 (with some tweaks) Atari-based 68k-systems running MiNT Microsoft Windows using Cygwin RISC OS ...just didn't make it back into mainline dillo1? 3. We could always borrow fltk's vsnprintf(), which is about 250 lines. I lean toward giving it an ifdef. -------------- Extra verbiage: Autoconf's manual says: The C99 standard says that if the output array isn't big enough and if no other errors occur, snprintf and vsnprintf truncate the output and return the number of bytes that ought to have been produced. Some older systems return the truncated length (e.g., GNU C Library 2.0.x or IRIX 6.5), some a negative value (e.g., earlier GNU C Library versions), and some the buffer length without truncation (e.g., 32-bit Solaris 7). Also, some buggy older systems ignore the length and overrun the buffer (e.g., 64-bit Solaris 7). glibc-2.1: 1999 solaris 7: 1998 irix 6.5: major release 1998 regular minor releases until 2005 "The last version of IRIX was 6.5.30, released in August 2006."

Sept. 16, 2014

dillo crash when dpid missing and view source used

by james.from.wellington＠gmail.com

Confirming this on OSX 10.6.8. Chmod a-x dipd is enough to exercise the fault. Here's a gdb backtrace where only html.cc is hacked, and it seems not to be relevant: --------------- (gdb) run Starting program: /Users/jamescone/buildTrees/dillo/versions/render/dillo-jfw01/src/dillo Reading symbols for shared libraries .+++++++++++++................................................................................ done Error in re-setting breakpoint 1: Function "main" not defined. paths: Cannot open file '/Users/jamescone/.dillo/keysrc': No such file or directory paths: Cannot open file '/usr/local/etc/dillo/keysrc': No such file or directory paths: Using internal defaults... Domain: Default accept. dillo_dns_init: Here we go! (threaded) Reading symbols for shared libraries . done Enabling cookies as from cookiesrc... Reading symbols for shared libraries . done ** WARNING **: preferred serif font "DejaVu Serif" not found. ** WARNING **: preferred monospace font "DejaVu Sans Mono" not found. ** WARNING **: preferred cursive font "URW Chancery L" not found. ** WARNING **: preferred fantasy font "DejaVu Sans" not found. Reading symbols for shared libraries . done Reading symbols for shared libraries . done Reading symbols for shared libraries . done Reading symbols for shared libraries . done Nav_open_url: new url='about:splash' a_Dicache_cleanup: length = 0 Nav_open_url: new url='dpi:/vsource/:about:splash' ** ERROR **: [Dpi_read_comm_keys] No such file or directory Dpi_start_dpid (child): Permission denied Dpi_start_dpid: can't start dpid ** ERROR **: dpi.c: can't start dpi daemon [dpid]: terminated normally (1) Program received signal EXC_BAD_ACCESS, Could not access memory. Reason: KERN_INVALID_ADDRESS at address: 0x0000000000000008 0x000000010001823d in a_Chain_check (FuncStr=0x1001013a0 "a_Capi_ccc", Op=2, Branch=1, Dir=2, Info=0x0) at chain.c:192 192 if (Info->Flags & (CCC_Ended + CCC_Aborted)) { (gdb) bt #0 0x000000010001823d in a_Chain_check (FuncStr=0x1001013a0 "a_Capi_ccc", Op=2, Branch=1, Dir=2, Info=0x0) at chain.c:192 #1 0x0000000100021137 in a_Capi_ccc (Op=2, Branch=1, Dir=2, Info=0x0, Data1=0x101914780, Data2=0x0) at capi.c:623 #2 0x0000000100021026 in a_Capi_dpi_send_data (url=0x0, bw=0x100550fa0, data=0x10192ee90 "<cmd='start_send_page' url='dpi:/vsource/:about:splash' data_size='9560' '>", data_sz=75, server=0x10010126b "vsource", flags=0) at capi.c:577 #3 0x00000001000210b3 in a_Capi_dpi_send_cmd (url=0x0, bw=0x100550fa0, cmd=0x10192ee90 "<cmd='start_send_page' url='dpi:/vsource/:about:splash' data_size='9560' '>", server=0x10010126b "vsource", flags=0) at capi.c:594 #4 0x00000001000208f9 in Capi_dpi_send_source (bw=0x100550fa0, url=0x100553e60) at capi.c:361 #5 0x0000000100020cd5 in a_Capi_open_url (web=0x100553e10, Call=0, CbData=0x0) at capi.c:440 #6 0x000000010001aad0 in Nav_open_url (bw=0x100550fa0, url=0x100551dd0, requester=0x0, offset=0) at nav.c:235 #7 0x000000010001ae74 in a_Nav_push (bw=0x100550fa0, url=0x100551dd0, requester=0x0) at nav.c:355 #8 0x000000010000b64e in UIcmd_open_url_nbw (new_bw=0x100550fa0, url=0x100551dd0) at uicmd.cc:761 #9 0x000000010000cfa8 in a_UIcmd_open_url_nt (vbw=0x101e1fe10, url=0x100551dd0, focus=1) at uicmd.cc:792 #10 0x000000010000d111 in a_UIcmd_view_page_source (bw=0x101e1fe10, url=0x101e0e680) at uicmd.cc:1214 #11 0x00000001000473fa in Menu_view_page_source_cb (user_data=0x101e0e680) at menu.cc:160 #12 0x000000010004774c in Fl_Menu_Item::do_callback (this=0x10018c7c0, o=0x10018c7c0) at Fl_Menu_Item.H:375 #13 0x0000000100046dbe in Menu_popup_cb (data=0x10018c790) at menu.cc:333 #14 0x00000001000a7db9 in do_timer () at object.cc:384 #15 0x00007fff84799bb8 in __CFRunLoopRun () #16 0x00007fff84797d8f in CFRunLoopRunSpecific () #17 0x00007fff868d374e in RunCurrentEventLoopInMode () #18 0x00007fff868d34b1 in ReceiveNextEventCommon () #19 0x00007fff868d340c in BlockUntilNextEventMatchingListInMode () #20 0x00007fff88f5deb2 in _DPSNextEvent () #21 0x00007fff88f5d801 in -[NSApplication nextEventMatchingMask:untilDate:inMode:dequeue:] () #22 0x00000001000a84ff in do_queued_events () at object.cc:384 #23 0x00000001000aabeb in fl_mac_flush_and_wait () at object.cc:384 #24 0x00000001000aeaa7 in Fl::run () at object.cc:384 #25 0x00000001000049f7 in main (argc=1, argv=0x7fff5fbff440) at dillo.cc:584 Current language: auto; currently c (gdb) --------- On 9/16/14, eocene <eocene at gmx.com> wrote: > > ...it happens at least as far back as 3.0.4 release. > > _______________________________________________ > Dillo-dev mailing list > Dillo-dev at dillo.org > http://lists.dillo.org/cgi-bin/mailman/listinfo/dillo-dev >

Sept. 16, 2014

dillo crash when dpid missing and view source used

by eocene＠gmx.com

...it happens at least as far back as 3.0.4 release.

Sept. 16, 2014

Patch: suppress duplicate Fl_Printer::class_id

by eocene＠gmx.com

James wrote: > Building commit 317f010 (two days ago) of dillo, I got a duplicate > symbol on OSX 10.6.8 with gcc 4.2.1. Here is the diff that suppresses > it. > > ---------------- > $ hg diff > diff -r c09473fb7311 dw/fltkplatform.cc > --- a/dw/fltkplatform.cc Sat Sep 13 09:46:28 2014 +0000 > +++ b/dw/fltkplatform.cc Sat Sep 13 23:49:26 2014 +1200 > @@ -37,10 +37,10 @@ > * any need for currently. This stub prevents that. */ > class FL_EXPORT Fl_Printer : public Fl_Paged_Device { > public: > - static const char *class_id; > +// static const char *class_id; > Fl_Printer(void) {}; > }; > -const char *Fl_Printer::class_id = "Fl_Printer"; > +//const char *Fl_Printer::class_id = "Fl_Printer"; To refresh my memory of the magnitude of what printer stuff we pull in without the stub, I commented out the stub -- and saw nothing to do with Fl_Printer with nm. I thought maybe this somehow had to do with the changes that Lauri Kasanen made to fltk last month like http://www.fltk.org/newsgroups.php?s10767+gfltk.commit+v10781 http://www.fltk.org/newsgroups.php?s10787+gfltk.commit+v10787 but I see that Manolo took out the Fl_Printer::class_id check in Fl_Text_Display in March 2012. 1.3.2 came out in...let's see... December 2012. So we should probably remove the stub altogether.

Sept. 15, 2014

Dillo and Irix

by eocene＠gmx.com

Back in April, I wrote: > I wrote: > > Jan wrote: > > > eocene wrote: > > > >Jan wrote: > > > >>Query: {GE} > > > >>a_IO_ccc : OpSend [1B] Info=10262f30 Flags=0 > > > > > > > >"GE" is the entire query? > > > > > > Yes exactly. Any idea? > > > > Can you watch what goes wrong in a_Http_make_query_str()? > > Between this and the file path thing, I wonder whether irix > doesn't like that va_list/va_copy/vsnprintf stuff in the dStr > code in dlib/dlib.c. It did turn out to be irix's idea of vsnprintf(). Modifying dStr_vsprintfa(), I browsed around for a few minutes without problems. Now to look into how to solve this properly...

Sept. 15, 2014

Implementing </div> closing <span>, and friends

by james.from.wellington＠gmail.com

Hello all, [Sorry if this is a dup; gmail may be playing tricks on me.] This page displays, with distortions, in dillo 3.0.4, but the image does not display after commit 317f010: https://medium.com/@verynicetweets/getting-out-of-the-house-is-so-hard-thes… A partially cleaned-up and stripped version is attached, where the comment block shows a suppression of the problem. The page is illegal: two unclosed <span>s interfere with a </div> and the following <div>. There is code around html.cc:1355, which is intended to implement heuristics to close unclosed tags but, in the cases where it is triggered, no tags are closed. Here is a patch which I think implements the intent of the heuristics: ----------------- $ hg diff diff -r 26378f85c4ea src/html.cc --- a/src/html.cc Sun Sep 14 09:56:22 2014 +0000 +++ b/src/html.cc Mon Sep 15 18:13:21 2014 +1200 @@ -1355,7 +1355,9 @@ i_SELECT = a_Html_tag_index("select"), i_TEXTAREA = a_Html_tag_index("textarea"); int w3c_mode = !prefs.w3c_plus_heuristics; - int stack_idx, tag_idx, matched = 0, expected = 0; + // TODO: If there was a clean way of never having a tag index zero, + // (it's currently <a>), then expected_idx could be prettier. + int stack_idx, tag_idx, matched = 0, expected_idx = -1, skipped = 0; TagInfo new_tag = Tags[new_idx]; /* Look for the candidate tag to close */ @@ -1368,26 +1370,42 @@ break; } else if (Tags[tag_idx].EndTag == 'O') { /* skip an optional tag */ + ++skipped; continue; } else if ((new_idx == i_BUTTON && html->InFlags & IN_BUTTON) || (new_idx == i_SELECT && html->InFlags & IN_SELECT) || (new_idx == i_TEXTAREA && html->InFlags & IN_TEXTAREA)) { /* let these elements close tags inside them */ + ++skipped; continue; } else if (w3c_mode || Tags[tag_idx].TagLevel >= new_tag.TagLevel) { /* this is the tag that should have been closed */ - expected = 1; - break; + /* skip it as if it were optional, but complain */ + ++skipped; + if (-1 == expected_idx) { + // ie this is the first thing we're unwinding + expected_idx = tag_idx; + } + continue; } } + // Either or both matched and expected_idx could now be set + + // TODO: decide whether to carry line numbers in the parser state stack, + // in order to report them here. + if (matched) { + if (-1 != expected_idx) { + BUG_MSG("Unexpected closing tag: </%s> -- expected </%s>", + new_tag.name, Tags[expected_idx].name); + } + Html_tag_cleanup_to_idx(html, stack_idx); - } else if (expected) { - BUG_MSG("Unexpected closing tag: </%s> -- expected </%s>.", - new_tag.name, Tags[tag_idx].name); } else { - BUG_MSG("Unexpected closing tag: </%s>.", new_tag.name); + BUG_MSG("Unexpected closing tag: </%s>" + " -- explored %d layer(s) and not closed.", + new_tag.name, skipped); } } ----------------- Closing tags that match no opening tag (I used </area>, which is illegal) do not cause catastrophic failure. So far, it has not broken on my usual workload. Gmail is noisy, so I wonder if something's going wrong there, but it looks fairly normal. I have no good idea what a general test suite for this code looks like. If someone wants to point me toward related tests, then I will think about extending them. Regards, James.

Sept. 15, 2014

dillo can't handle strings

by mail＠gabucino.hu

On Sep 12, 2014, at 5:18 AM, Gabor Berczi wrote: > > On Sep 11, 2014, at 11:09 PM, eocene wrote: > >> I'd be interested in either, although if the build/debug process >> has some >> important differences from what a linux user would be accustomed >> to, I >> might need a quick bit of orientation... > > I've sent you the login credentials. As for IRIX only Jan can help, > since my Octane's ethernet interface developed a buffer RAM fault. eocene couldn't figure out how to use remote X instead of Quartz (he knows only linux), any other takers? -- G

Sept. 14, 2014