April 2009 - Darwin-dev - lists.apple.com

testing the bootstrap context for validity?
by Bill Janssen 02 Apr '09

02 Apr '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com [I mistakenly sent this to darwin-kernel...] I seem to have a situation where the bootstrap context of a long-running daemon process sometimes "goes bad". Not sure just what's happening, but it can't seem to use any Mach ports. Can't connect to the windowServer, can't do DNS lookups, etc. When I use 'launchctl bslist' to look at the BC, I get an odd output: > % sudo launchctl bslist 7942 > bootstrap_info(): 268435459 > % Is there a function I could call from a process to check the current bootstrap context? If the process could detect, in some systematic way, whether or not its bootstrap context is still good, it could restart when it needs to. (I can't call launchctl, because the process doesn't run as root.) Better yet, is there some way to prevent this context decay? My daemon is not launched via launchd, it's just classic UNIX daemonization: fork() setsid() fork() exec() Bill _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Re: sending password from pre-efi boot prompt to iokit driver
by David Elliott 02 Apr '09

02 Apr '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Apr 1, 2009, at 1:15 PM, Joel Reymont wrote: -Dave _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… What mechanism should I use to share a password entered at the EFI pre-boot prompt to a driver? I thought of storing the password in NVRAM but don't believe it would go away upon reboot. It's imperative that the password evaporates once a machine is rebooted. It's been a while since I looked at this but I'm almost certain that the EFI variable protocol has the ability to store variables either in NVRAM or in a temporary store. When you set the value of a variable you specify where you want it to be stored. So that probably will work. You should investigate it further, perhaps by looking at the Tiano code. Failing that, you can always produce a configuration table allocated out of runtime RAM. The booter will examine all of the entries in the system table's configuration table list and make the address of the table data available as the table property of an entry in IODeviceTree:/efi/configuration-table whose name is the GUID as a string. Or at least I _think_ the booter (boot.efi) just looks at all the configuration tables listed by the system table and outputs their pointers into the device tree. I can't imagine it being coded to filter the list. Obviously without source it's impossible to know, but this would be my guess. This email sent to site_archiver(a)lists.apple.com

1 0

Are kqueue events "round robin"?
by Påhl Melin 02 Apr '09

02 Apr '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:received:date:message-id:subject :from:to:content-type:content-transfer-encoding; bh=xPgnTjCzi+BRbw/NeGhgy7q9O7bm9arTD2vJ8FmnAC8=; b=Fj9NAFyQHzWGWDhS/Uql2N2PnhhS5l61W4X/+Gnk3yv6sX0IopinLqoOi/IptySENn 4NVpwRuEGcJsoD2mcUv/FXE0KseU5qrIP/eYdY93HpRCKZBYbEIi1rWIMnnUkK/JA8fv pRguf6lxjgwpJIkMUQUKCG+byYVuUt+ePM2sI= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:date:message-id:subject:from:to:content-type :content-transfer-encoding; b=yHBjsgSA0AI136/MnCajfqoyD3q1RfTRam1lDG1exRkk4tPZDb2O6T2ABnZ1t92gdR Z4vl+dGtBzLUnz9ngwW2BIJc1TFIlXOzSCbQ1F4Hl6A7rXh9H/j97EU3yx78m5wQkaqs 4D752xQZeVK3vYVHM8lsPkeCiZO5ouYEaT5DI= Hi all, I have a question regarding the order events are returned from a kevent() call on a kqueue, if I only supply one event in the event list and no events are changed? Is there a predefined order in which the filters are processed and returned as events when there are multiple events available at the same time? For example if I have two filters, one EVFILT_READ filter and one EVFILT_WRITE filter and there is available room in both the read and the write buffers. If I make four subsequent calls, in what order will the events be returned? Will I get EVFILT_READ, EVFILT_WRITE, EVFILT_READ, EVFILT_WRITE? Or will I e.g. get one of the events repeatedly until that buffer is empty/full, like EVFILT_READ until I have read everything in the read buffer? Or is random or some other algorithm? What I want to do is to alternate between read and write events to get good responsiveness. I suppose I could change the events in the kqueue but it would be easier to rely on some guaranteed behavior. Thanks in advance, / Påhl _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Why DYLD_INSERT_LIBRARIES works differently on 10.4 and 10.5
by Alexander Samoylovich 01 Apr '09

01 Apr '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; q=dns/txt; l=2071; t=1238623657; x=1239487657; c=relaxed/simple; s=sjdkim2002; h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version; d=cisco.com; i=asamoylo(a)cisco.com; z=From:=20Alexander=20Samoylovich=20<asamoylo(a)cisco.com> |Subject:=20Why=20DYLD_INSERT_LIBRARIES=20works=20different ly=20on=2010.4=20and=2010.5 |Sender:=20; bh=PDvBpT1JRg41OsANSUuplPljhab+3jYAR/eqpDJomak=; b=YBY+B42ZY9jDQwRMgWSINl/5dQ5nm4CdC7a7tSCnXZdVaR/7fog5X8q6dA etoXw6efeCbcIk1NZ/Iz+Lt1tcLKg9IC+wOweqD1+uxwQMzF+LXDRt3sjNw7 BYOvI8H+xF; Hello. So the question: Thank you, Alex hook.dylib is the library to be inserted: #include <stdio.h> #include <sys/socket.h> #include <sys/types.h> #include <sys/uio.h> #include <netdb.h> #include <unistd.h> #include <stdarg.h> typedef struct interpose_s { void* new_func; void* old_func; } interpose_t; void report(char *str) { FILE *f = fopen("hook.log", "a"); fprintf(f, "%s", str); fclose(f); } void my_freeaddrinfo(struct addrinfo *ai) { report("--freeaddrinfo--\n"); freeaddrinfo(ai); } I am trying to interpose getaddrinfo and freeaddrinfo functions. For this purpose I use DYLD_INSERT_LIBRARIES environment variable. The library source and the shell script to run an application to be hooked are below. This trick works for Mac OS X 10.4 for Safari, Firefox and curl. But under Mac OS X 10.5 it works with Firefox and curl but not with Safari. For Safari the library hooks freeaddrinfo but not getarrdinfo. The real application also interposes connect and connect$UNIX2003 and works for all application and all OSes. Why DYLD_INSERT_LIBRARIES works differently on 10.4 and 10.5? Is there any way to force it to work the same way and to interpose getaddrinfo too? I use the following shell script to start a process, Safari in this case: DYLD_INSERT_LIBRARIES=hook.dylib /Applications/Safari.app/Contents/ MacOS/Safari extern int my_getaddrinfo(const char * __restrict, const char * __restrict, const struct addrinfo * __restrict, struct addrinfo ** __restrict); extern void my_freeaddrinfo (struct addrinfo *ai); int my_getaddrinfo(const char * __restrict a, const char * __restrict b, const struct addrinfo * __restrict c, struct addrinfo ** __restrict d) { report("--getaddrinfo--\n"); return getaddrinfo (a, b, c, d); } const interpose_t interposes[] __attribute__ ((section ("__DATA, __interpose"))) = { {(void*) my_getaddrinfo, (void*) getaddrinfo}, {(void*) my_freeaddrinfo, (void*) freeaddrinfo}}; _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

sending password from pre-efi boot prompt to iokit driver
by Joel Reymont 01 Apr '09

01 Apr '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:received:received:message-id:from:to :content-type:content-transfer-encoding:mime-version:subject:date :x-mailer; bh=/bVkh8uU3TsH1sJsReo4uChuYfSnSy2wivf7+hdb+LU=; b=SxCNlAkwPE7sA0Mfo7RQf0prSwWtIiHV81DxI6lTkufnYoJXapg3EmR4RRfmnhgP+t sLgrOKz3XfAXLohjMJY3UrWdEfJ4GJZqaElTVnPgAl8qq+KOrI6CvhYTXKtFzO16o0mG hMtnsQ5bCfHU23TjGr49M9x83Y2PmEsq/Cv2Q= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:from:to:content-type:content-transfer-encoding :mime-version:subject:date:x-mailer; b=xFvbJPhk4lkGe+PQCLqaC3j3t4En3CHNNMqNIdct6m+c9IS3283sSaZwi9+C7d0EMj EUMmz14z0wySDX9Cwdf7PdOXlmprOUKNOUR80+FjnWx4Ha0HWUZWMG9b6Brm09kLYz/s luUjPWAYmXULhP35HDPq5O+Q2UUKjH2F9EPSM= Any suggestions? Thanks, Joel --- http://tinyco.de Mac, Lisp, OCaml _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… What mechanism should I use to share a password entered at the EFI pre- boot prompt to a driver? I thought of storing the password in NVRAM but don't believe it would go away upon reboot. It's imperative that the password evaporates once a machine is rebooted. This email sent to site_archiver(a)lists.apple.com

1 0