- Dillo-dev - lists.mailman3.com

Re: [Dillo-dev]pthreads

by Livio Baldini Soares

Hi (again), Livio Baldini Soares writes: > Hi Jorge, > > I just got the chance to look at your little test program. It has a > few bugs: > > 1) In pthread_create() you are passing a pointer to the unintialized > thrdID[i] variable. Pthreads does not guarantee that the threadID > argument is initialized before the start_routine is called. If you > try printing the value of p_thrID in the 'thr_function', you'll see > you'll get some correct and some random values. I just looked and there are three occurences of this bug in Dillo: dpi/file.c src/IO/IO.c src/dns.c All three of them should have the current calls to: 'pthread_detach(uninitialized-variable)' changed to use 'pthread_detach(phtread_self())' As in all three cases the variable used has some likelihood (in my system, a high likelihood) of not yet being initialized. In some cases (like struct DnsServer) the threadID variable can optionally be altogether removed. The final patch is left as an exercise to the reader ;-) hope this helps, -- Livio B. Soares

Nov. 25, 2004

Re: [Dillo-dev]pthreads

by robert w hall

In article <20041125152322.GH1676(a)dillo.org>, Jorge Arellano Cid <jcid(a)dillo.org> writes > What an irony: after years of chasing the most minor memory >leaks (up to a few bytes), it ends up being that we've leaked >a lot, by Megabyte units, because of an external library. > well I _did_ say it didn't appear to be leaking on a playstation-2... -- robert w hall

Nov. 25, 2004

Re: [Dillo-dev]pthreads

by Livio Baldini Soares

Hi Jorge, I just got the chance to look at your little test program. It has a few bugs: 1) In pthread_create() you are passing a pointer to the unintialized thrdID[i] variable. Pthreads does not guarantee that the threadID argument is initialized before the start_routine is called. If you try printing the value of p_thrID in the 'thr_function', you'll see you'll get some correct and some random values. 2) Still, it's not a good idea to pass variables from the thread stack. In this particular case, it will work since the main thread outlives all the created ones, but it is usually highly likely to cause you headaches... A minimal fix for your test code is: --- pth_mem.orig.c 2004-11-25 11:16:01.000000000 -0500 +++ pth_mem.c 2004-11-25 11:33:34.000000000 -0500 @@ -7,11 +7,10 @@ int mode; static void *thr_function(void *data) { int i; - pthread_t *p_thrID = data; void *mem; if (mode == 'd') - pthread_detach(*p_thrID); + pthread_detach(pthread_self()); for (i = 0; i < 10; ++i) { mem = malloc(1024*1024); In my box, this gets rid of any leaks. best regards, -- Livio B. Soares

Nov. 25, 2004

Re: [Dillo-dev]pthreads

by Lars Segerlund

I REALLY understand chasing the leaks and then finding something like this stings ... / regars, Lars Segerlund. On Thu, 25 Nov 2004 12:23:22 -0300 Jorge Arellano Cid <jcid(a)dillo.org> wrote: > Hi, > > Well, I searched for where to report bugs in pthreads and the > process is tedious to say the least. > > They want the submitter to test against CVS glibc which is a > thing I don't have the time, nor the machine to do. > > FWIW, a near target is at least the test program on the latest > glibc (glibc-2.3.3 at this time), but I have 2.3.2. Can someone > please test it with glibc-2.3.3? > > If the Solaris test shows no leaks, then that's strong evidence > of a _serious_ problem within glibc's pthreads and the bug report > will most probably be welcomed. > > After I have the results from glibc-2.3.3 and Solaris, I'll > send the report. In the meanwhile I've been playing some > workarounds for Dillo and the results is an impressive memory > usage reduction so far (BTW, the reload memory leak also had its > roots in pthreads leaking memory). > > What an irony: after years of chasing the most minor memory > leaks (up to a few bytes), it ends up being that we've leaked > a lot, by Megabyte units, because of an external library. > > Notwithstanding there's still the possibility of me being wrong > in the understanding of its API, this chance is growing thiner. > > It could also be the kernel (or ps) not reporting correctly the > VSZ and RSS. Who knows? At least the workarounds are improving a > lot these stats in my tree. > > > -- > Cheers > Jorge.- > > _______________________________________________ > Dillo-dev mailing list > Dillo-dev(a)dillo.org > http://lists.auriga.wearlab.de/cgi-bin/mailman/listinfo/dillo-dev

Nov. 25, 2004

Re: [Dillo-dev]pthreads

by Jorge Arellano Cid

Hi, Well, I searched for where to report bugs in pthreads and the process is tedious to say the least. They want the submitter to test against CVS glibc which is a thing I don't have the time, nor the machine to do. FWIW, a near target is at least the test program on the latest glibc (glibc-2.3.3 at this time), but I have 2.3.2. Can someone please test it with glibc-2.3.3? If the Solaris test shows no leaks, then that's strong evidence of a _serious_ problem within glibc's pthreads and the bug report will most probably be welcomed. After I have the results from glibc-2.3.3 and Solaris, I'll send the report. In the meanwhile I've been playing some workarounds for Dillo and the results is an impressive memory usage reduction so far (BTW, the reload memory leak also had its roots in pthreads leaking memory). What an irony: after years of chasing the most minor memory leaks (up to a few bytes), it ends up being that we've leaked a lot, by Megabyte units, because of an external library. Notwithstanding there's still the possibility of me being wrong in the understanding of its API, this chance is growing thiner. It could also be the kernel (or ps) not reporting correctly the VSZ and RSS. Who knows? At least the workarounds are improving a lot these stats in my tree. -- Cheers Jorge.-

Nov. 25, 2004

Re: [Dillo-dev]Ugly fonts

by Livio Baldini Soares

Hi, Sorry for the slightly unrelated post, higuita writes: > On Wed, 24 Nov 2004 17:57:19 -0500 (EST), "Chris de Vidal" <Chris(a)deVidal.tv> wrote: > > Switched vw_fontname to > > "-misc-fixed-bold-r-normal--12-120-75-75-c-0-iso8859-1" > [...] > > kernel version, but running KDE 3.3 not XFCE. Looks beatiful, just like > > the screenshot. > > probably kde installed truetype fonts... > try to do the xfontsel in that computer and compare > then You probably mean 'xlsfonts', which will list all the fonts the current X server has access to (in a nicer format for comparison, than xfontsel). Chris, regarding your Debian box, make sure you have correctly configured your '/etc/fonts/local.conf' configuration file. Specifically, I have found that most Debian default installations disable bitmapped fonts, due to the use of anti-aliased or truetype fonts used in most applications and/or window managers. Uncomment the lines around: <dir>/usr/X11R6/lib/X11/fonts</dir> After you do that, run: $ defoma-reconfigure and $ fc-cache -f Another issue is that fonts with "odd" sizes are butt-ugly. In this respect, try adding 0.1 to your "font_factor" variable in your dillorc file (if it's 1.0 set it to 1.1, if it's 1.1 set it to 1.2, for example). It usually makes very little difference in the final font size, but makes it a _big_ difference in scalability. There was something else I was going to suggest, but I forget now... If I remember I'll write. Sorry, have to run now, good luck, and hope this helps, -- Livio B. Soares

Nov. 25, 2004

Re: [Dillo-dev]Ugly fonts

by higuita

On Wed, 24 Nov 2004 17:57:19 -0500 (EST), "Chris de Vidal" <Chris(a)deVidal.tv> wrote: > Switched vw_fontname to > "-misc-fixed-bold-r-normal--12-120-75-75-c-0-iso8859-1" try "-misc-fixed-*-*-*--*-*-*-*-*-*-iso8859-1" maybe this way you manage to get the bold, italic and different font sizes the way you put force the font to be exactly like that, my way "might" allow all combinations (might because i'm not sure if this isnt too generic or if dillo support it) also, is this the only font you have in xfontsel? > the dillo.org screenshot has very smooth fonts. Mine is all jaggedy. the jagged font usually is because the font is not propocional/scalable or low quality > kernel version, but running KDE 3.3 not XFCE. Looks beatiful, just like > the screenshot. probably kde installed truetype fonts... try to do the xfontsel in that computer and compare then good luck higuita -- Naturally the common people don't want war... but after all it is the leaders of a country who determine the policy, and it is always a simple matter to drag the people along, whether it is a democracy, or a fascist dictatorship, or a parliament, or a communist dictatorship. Voice or no voice, the people can always be brought to the bidding of the leaders. That is easy. All you have to do is tell them they are being attacked, and denounce the pacifists for lack of patriotism and exposing the country to danger. It works the same in every country. -- Hermann Goering, Nazi and war criminal, 1883-1946

Nov. 25, 2004

Re: [Dillo-dev]Ugly fonts

by Jorge Arellano Cid

On Wed, Nov 24, 2004 at 05:57:19PM -0500, Chris de Vidal wrote: > Thus spaketh higuita: > > check with xfontsel what font do you really have for normal > > X programs, as gtk2 programs should be using a new way of > > generating fonts (XFT) and probably most of your truetype > > fonts arent really in the old X11 font interface > > > > try chossing a font listed in the xfontsel that looks > > good and its scalable... > > if you have bitstream vera, use it > > Switched vw_fontname to > "-misc-fixed-bold-r-normal--12-120-75-75-c-0-iso8859-1" > which looks somewhat better, but still not as good as the originals: > http://devidal.tv/~chris/mine_versus_dillo.org.gif > > You may not see much difference but believe me my eyes hurt looking at my > dillo. You really see the difference in the two green topic title bars; > the dillo.org screenshot has very smooth fonts. Mine is all jaggedy. > > Tried the same version of dillo on my wife's PC, also Debian Sid, same > kernel version, but running KDE 3.3 not XFCE. Looks beatiful, just like > the screenshot. > > Something isn't right. I thought I set her machine up with the same fonts... > > Ideas anyone? Hi. It should be working. This is a problem with your machine's installed fonts, not with Dillo. Dillo doesn't do True Type at this time, so there's no point installing this type of fonts. I'm not a Debian user, but most probably you left out a set of "legacy" fonts that Dillo requires. Maybe the fastest solution is to check the font-specific .debs of your wife's cpmputer against yours, once this is done update the font cache, restart X and retry. A more accurate way of solving it is to ask a long time Debian user what the packages for "legacy" (not antialiased) fonts are and how to install them. Please let us know when it starts working for you. HTH. -- Cheers Jorge.-

Nov. 24, 2004

Re: [Dillo-dev]Ugly fonts

by Chris de Vidal

Thus spaketh higuita: > check with xfontsel what font do you really have for normal > X programs, as gtk2 programs should be using a new way of > generating fonts (XFT) and probably most of your truetype > fonts arent really in the old X11 font interface > > try chossing a font listed in the xfontsel that looks > good and its scalable... > if you have bitstream vera, use it Switched vw_fontname to "-misc-fixed-bold-r-normal--12-120-75-75-c-0-iso8859-1" which looks somewhat better, but still not as good as the originals: http://devidal.tv/~chris/mine_versus_dillo.org.gif You may not see much difference but believe me my eyes hurt looking at my dillo. You really see the difference in the two green topic title bars; the dillo.org screenshot has very smooth fonts. Mine is all jaggedy. Tried the same version of dillo on my wife's PC, also Debian Sid, same kernel version, but running KDE 3.3 not XFCE. Looks beatiful, just like the screenshot. Something isn't right. I thought I set her machine up with the same fonts... Ideas anyone? CD "You will exist forever. You and God are both in the universe to stay either as friends on his terms, or enemies on yours. Which it will be is proven in this life. And this life is a vapor. Two seconds, and we will be gone. Surely God means for our time on earth to count for something significant." -- John Piper in "Life as a Vapor." Read it here: http://www.desiringgod.org/library/onlinebooks_index.html#blv

Nov. 24, 2004

Re: [Dillo-dev]pthreads

by Jorge Arellano Cid

On Wed, Nov 24, 2004 at 03:15:15PM +0100, Riccardo Mottola wrote: > Hey Jorge, > > just a quick answer, but I am ina hurry > > > I have solaris, I have gcc. so I can test and compile. I have also > MkLinux which is linux where threads are mapped to mach threads (linux > on mach microkernel). > > > What source code should I run? How do i check for memory leaks (no > valgrind for solaris)... I posted it to the mailing list with instructions. You can find it here: http://lists.auriga.wearlab.de/pipermail/dillo-dev/2004-November/002436.html The mail thread starts at: http://lists.auriga.wearlab.de/pipermail/dillo-dev/2004-November/002429.html BTW, I'm making some tests in dillo using the results explained there, and the results are impressive: For instance: USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND jcid 3234 2.5 13.9 663804 13116 pts/15 S+ 11:59 0:05 ./dillo jcid 3238 0.0 13.9 663804 13116 pts/15 S+ 11:59 0:00 ./dillo (see the huge virtual size? It's a simple page with lots of tiny images). The same dillo, this time using the explicit detached-attribute pthread creation technique, gives: USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND jcid 3792 11.8 11.0 14116 10448 pts/15 S+ 12:49 0:07 ./dillo jcid 3793 0.3 11.0 14116 10448 pts/15 S+ 12:49 0:00 ./dillo Same code, same page. Something really weird is going on here! Does anyone know the email of the pthreads library maintainer? -- Cheers Jorge.-

Nov. 24, 2004