+2002-12-26 Martin Mares <mj@ucw.cz>
+
+ * Added preliminary version of NetBSD support by Quentin Garnier
+ <netbsd@quatriemek.com>.
+
+2002-04-06 Martin Mares <mj@ucw.cz>
+
+ * lspci.c: Mention "-xxx" in the help.
+
+2002-03-30 Martin Mares <mj@ucw.cz>
+
+ * Released as 2.1.10.
+
+ * Removed pci.ids from the CVS as they are now maintained separately.
+ They are still included in the tarball, but if you are building pciutils
+ from CVS sources or you want to get a newer ID file than the one distributed,
+ just use "make update-ids" to download the most recent snapshot from
+ pciids.sf.net.
+
+ * Makefile, README: Updated accordingly.
+
+ * Makefile: Added a couple of maintenance targets.
+
+ * maint/*: Maintainer's utilities now reside here (present only in CVS,
+ not in the tarball).
+
+ * lib/names.c (parse_name_list): End-of-line comments are no longer
+ supported. Hashes are now perfectly valid in all names and they start
+ a comment only at the start of the line.
+
+2002-03-26 Martin Mares <mj@ucw.cz>
+
+ * README: Rewritten.
+
+ * Makefile: When the currently installed pci.ids file is newer than
+ the version to be installed, don't overwrite it. Suggested by
+ Jean Delvare <khali@linux-fr.org>.
+
+2002-03-24 Martin Mares <mj@ucw.cz>
+
+ * lspci.c (show_pcix_bridge): Added PCI-X capability display.
+ Contributed by Jeff Garzik <jgarzik@mandrakesoft.com>.
+
+ * lib/header.h: Added PCI-X stuff. Again by Jeff.
+
+ * Makefile (CFLAGS): -Werror is gone.
+
+ * lspci.c (format_agp_rate): Fixed AGP rate reporting.
+
+ * lib/filter.c (pci_filter_parse_slot): Allow bus 0xff, slot 0x1f and
+ function 7.
+
+ * lib/names.c (pci_lookup_name): When printing unknown subsystem vendor
+ or device hexadecimally, don't confuse it with chip vendor/device ID.
+ First reported by Marc Boucher <marc@mbsi.ca>.
+
+2001-11-04 Martin Mares <mj@ucw.cz>
+
+ * Released as 2.1.9.
+
+ * pci.ids: Synced with the PCI ID Repository at SourceForge
+ (http://pciids.sourceforge.net/).
+
+2000-08-25 Martin Mares <mj@albireo.ucw.cz>
+
+ * pci.ids: Updated ID's, synced with kernel 2.4.0-test7.
+
+2000-06-13 Martin Mares <mj@albireo.ucw.cz>
+
+ * Makefile (uninstall): Fixed.
+
+ * pci.ids: ID updates.
+
+2000-05-20 Martin Mares <mj@albireo.ucw.cz>
+
+ * Released as 2.1.8.
+
+ * pci.ids: ID updates.
+
+ * lspci.c (show_verbose): Min/max latency is now printed in nanoseconds.
+
+2000-05-04 Martin Mares <mj@albireo.ucw.cz>
+
+ * lspci.c (show_verbose): If the header type doesn't match device
+ class, just print a warning and still use the header type, assuming
+ bogus class info.
+
+2000-05-01 Martin Mares <mj@albireo.ucw.cz>
+
+ * lspci.c (show_pm): Updated according to PCI PM specs 1.1.
+
+ * pci.ids: Updated Contemporary Control ARCnet card entries.
+
+ * pci.ids: Synchronized class list with latest PCI SIG documents.
+
+2000-04-29 Martin Mares <mj@albireo.ucw.cz>
+
+ * lib/names.c (pci_lookup_name): Fixed the cheat.
+
+ * lib/internal.h: Moved UNUSED/inline hacks above all other declarations.
+
+ * pci.ids: Various ID updates.
+
+2000-04-21 Martin Mares <mj@albireo.ucw.cz>
+
+ * Released as 2.1.7.
+
+ * pciutils.spec: Added German description.
+
+ * lib/names.c (pci_lookup_name): Added a cheat for case that subsystem ID's
+ are the same as device ID and just output subsystem name according to
+ device name.
+
+ * lib/proc.c: Use UNUSED instead of __attribute__((unused)).
+
+ * lib/pci.h: Added the correct typedefs for AIX.
+
+ * lib/generic.c (pci_generic_scan_bus): Export.
+
+ * lib/configure: Detect AIX. Use "echo '...\c'" instead of
+ "echo -n" if the shell isn't bash.
+
+ * lspci.c: Avoid using alloca() if not compiled by GCC.
+ (insert_dev): Avoid returning a void value.
+
+ * Makefile: Choose which install utility to use according to
+ OS type. Also use `-o' instead of `-or' in find arguments.
+ Part of the AIX port.
+
+ * lib/aix-device.c: Added AIX low-level interface. Contributed
+ by Dave Johnson <ddj@brown.edu>, thanks a lot!
+
+ * pci.ids: Few new ID's.
+
+ * pciutils.h, lib/internal.h: Redefine __attribute__ and inline
+ when not compiled by GCC.
+
+2000-04-17 Martin Mares <mj@albireo.ucw.cz>
+
+ * Released as 2.1.6.
+
+ * pci.ids: Added a large batch of ID's contributed by Gunther
+ Mayer <gunther.mayer@braunschweig.netsurf.de> and synced the
+ list with the 2.3.99pre6 kernel.
+
+Thu Mar 9 13:11:39 2000 Martin Mares <mj@albireo.ucw.cz>
+
+ * lib/example.c: Added a simple example of how to use
+ the library.
+
+ * lspci.man, setpci.man: Revealed --version. Well spotted
+ by Adam Sulmicki.
+
+ * pci.ids: Added lots of new ID's and synced with 2.3.50.
+
+Tue Feb 15 00:16:21 2000 Martin Mares <mj@albireo.ucw.cz>
+
+ * Released as 2.1.5.
+
+ * pciutils.spec: Updated spec file from Andreas Jaeger. Now using
+ %{_mandir} to find the right location to install manual pages to.
+
+ * pci.ids: New ID's.
+
+ * setpci.c (main): Don't crash when `-s' or `-d' is given with
+ no argument.
+
+Mon Jan 24 20:04:11 2000 Andreas Jaeger <aj@suse.de>
+
+ * lib/i386-ports.c: Include <asm/io.h> only on systems without
+ glibc.
+
+Mon Jan 24 11:24:43 2000 Martin Mares <mj@albireo.ucw.cz>
+
+ * Makefile: Modified to work with the Linux FHS brain-damaged idea
+ of putting man pages to /usr/share/man. Now we choose either /usr/man
+ or /usr/share/man depending on which one exists.
+
+Sun Jan 23 10:52:34 2000 Martin Mares <mj@albireo.ucw.cz>
+
+ * pci.ids: Synced with Linux 2.3.40.
+
+ * Released as version 2.1.4.
+
+Thu Jan 20 11:08:32 2000 Martin Mares <mj@albireo.ucw.cz>
+
+ * lspci.c (scan_device): Fixed bug in reading of cardbus extensions.
+
+ * lib/proc.c: pread/pwrite fixed again, this time on libc5 :(
+
+ * lspci.c (format_agp_rate): Better formatting of AGP rates.
+
+ * pci.ids: New ID's.
+
+ * lib/configure: Added configuration for ia64 (the same as for Alpha).
+ Patch from Stephane Eranian <eranian@cello.hpl.hp.com>.
+
+Thu Jan 13 23:01:52 2000 Martin Mares <mj@albireo.ucw.cz>
+
+ * Released as version 2.1.3.
+
+ * pci.ids: New ID's.
+
+ * lspci.c: Alpha IO_FORMAT fix by Andrea Arcangeli.
+
+ * lib/access.c: Corrected order of access methods -- it's wonderful how
+ long could such a bug remain unspotted. Thanks go to Andreas Schockenhoff.
+
+Sat Dec 11 23:34:48 1999 Martin Mares <mj@albireo.ucw.cz>
+
+ * Released as version 2.1.2.
+
+ * lib/names.c (pci_lookup_name): Fixed printing of unknown subsystem ID's.
+
+Sat Dec 4 13:11:03 1999 Martin Mares <mj@albireo.ucw.cz>
+
+ * Released as version 2.1.1.
+
+ * pci.ids: Several new ID's.
+
+ * setpci.c (exec_op): Don't forget to increment address after writing
+ single value, so that `30.b=12,34,56,78' works as expected. Reported
+ by Petr Soucek <petr@ryston.cz>.
+
+Wed Nov 3 10:24:38 1999 Martin Mares <mj@albireo.ucw.cz>
+
+ * lspci.c (show_terse): If prog-if is zero, but it's present in the
+ ID list, show it.
+
+Sat Oct 30 11:06:35 1999 Martin Mares <mj@albireo.ucw.cz>
+
+ * Released as version 2.1.
+
+ * pci.ids: Synced to devlist.h in 2.3.24 kernel. Merged in
+ lots of contributed entries.
+
+Thu Oct 14 10:40:12 1999 Dave Jones <dave@powertweak.com>
+
+ * pci.ids: Added lots of subsystem definitions.
+
+Sat Oct 9 14:32:28 1999 Martin Mares <mj@albireo.ucw.cz>
+
+ * setpci.man: Better example.
+
+ * lspci.c: Resolve prog-if names.
+
+ * lib/names.c: Adapted to new pci.ids syntax.
+
+ * pci.ids: Reorganized the pci.ids file. Subsystems are listed
+ under devices, classes can optionally contain prog-if information.
+ New ID's, as usually.
+
+Wed Sep 22 09:45:24 1999 Martin Mares <mj@albireo.ucw.cz>
+
+ * pci.ids: New ID's. As usually.
+
+ * lspci.c (show_htype1, show_htype2): Don't show `closed' bridge windows
+ unless we're vvvery verbose.
+
+Mon Sep 20 11:22:54 1999 Martin Mares <mj@albireo.ucw.cz>
+
+ * lspci.c (show_htype2): Don't forget a TAB before the "I/O window" line.
+
+Tue Sep 14 09:31:01 1999 Martin Mares <mj@albireo.ucw.cz>
+
+ * pci.ids: New ID's. Again.
+
+Thu Aug 5 17:45:42 1999 Martin Mares <mj@albireo.ucw.cz>
+
+ * lspci.c (show_size): Don't show size if it's zero (usually
+ due to bus-centric mode being enabled).
+
+Tue Jul 20 13:25:09 1999 Martin Mares <mj@albireo.ucw.cz>
+
+ * lib/pci.h: Types are OS dependent.
+
+ * lib/internal.h: Byte order hacks are OS dependent.
+
+ * lib/configure: Recognize FreeBSD. Also set OS_XXX according to the OS.
+
+ * lib/Makefile, lib/access.c: Added fbsd-device target.
+
+ * pci.ids: Added an entry for PLX 9080 bridges with subsystem ID
+ incorrectly set to device ID.
+
+ * README: Added a note that we support FreeBSD.
+
+ * Makefile (PREFIX, ROOT): Set it depending on the OS.
+ (%.8): Grrr, BSD date is not able to convert date formats the same way
+ as GNU date does. Use sed instead.
+ (install): Use `-c' when calling install.
+
+ * lib/fbsd-device.c: Added FreeBSD /dev/pci access module contributed
+ by Jari Kirma <kirma@cs.hut.fi>.
+
+ * lib/proc.c: Rewrote the pread/pwrite things once again. Use pread
+ and pwrite only when we are certain it's safe (i.e., glibc 2.1
+ on all architectures or any libc on a i386 where we really know
+ how to use syscalls directly). In all other cases, emulate it
+ with lseek/read/write.
+
+ * pci.ids: Some more IDs.
+
+Mon Jul 19 14:10:36 1999 Martin Mares <mj@albireo.ucw.cz>
+
+ * pci.ids: Added IDs of parallel port controller cards as sent
+ by Tim Waugh <tim@cyberelk.demon.co.uk>.
+
+Wed Jul 7 00:55:48 1999 Martin Mares <mj@albireo.ucw.cz>
+
+ * lib/proc.c (proc_scan): HAVE_LONG_ADDRESS instead of HAVE_64BIT_ADDRESS.
+
+ * lspci.c: ADDR_FORMAT now depends on both HAVE_64BIT_ADDRESS and
+ HAVE_LONG_ADDRESS.
+
+ * lib/configure: HAVE_64BIT_ADDRESS now doesn't imply that the addresses
+ are represented as long long. Introduced a new flag HAVE_LONG_ADDRESS
+ to indicate this. Both Sparc architectures now use both 64BIT_ADDRESS
+ and LONG_ADDRESS since they have 32-bit userland, but don't set LONG_ADDRESS
+ on Alpha.
+
+ * lspci.c (show_msi): Added dumping of the MSI capability.
+ (show_slotid): The same for SlotID capability.
+ (show_caps): Seperated capability dumping, because it should
+ be used for both htype0 and htype1. Even PCI 2.2 doesn't mention
+ layout of htype2, so I'm a bit unsure about it wrt capabilities
+ -- they at least have to live somewhere else since address 0x34
+ is already defined differently in htype2.
+
+ * lib/header.h (PCI_STATUS_UDF): Marked as obsolete (PCI 2.2).
+ (PCI_BASE_ADDRESS_MEM_TYPE_1M): Ditto.
+ (PCI_CAP_ID_*): Added some new capabilities defined in PCI 2.2.
+
+ * lspci.c (show_htype0): Mask out lowest 2 bits from all capability pointers
+ (required by the PCI 2.2 specs).
+
+Mon Jul 5 12:45:19 1999 Martin Mares <mj@albireo.ucw.cz>
+
+ * lspci.c (show_size): Added new function for dumping of region sizes.
+ (show_rom): Added. Now correctly dumps unassigned ROM addresses, disabled
+ ROMs et cetera.
+
+ * lib/pci.h (struct pci_dev): known_fields moved to the public part
+ of the structure.
+
+ * Added support for region sizes. Needs support in the kernel, because
+ sizes have to be scanned before drivers initialize yourself. pci_fill_info()
+ now returns a bitmask of what fields were really read.
+
+Mon Jun 28 18:37:51 1999 Dave Jones <dave@powertweak.com>
+
+ * Makefile:
+ Added `uninstall' target - removes all traces of pciutils.
+ * pci.ids:
+ - Lots of cleanups in the ATI, S3, Adaptec vendors.
+ - Changed Compaq QVision 1280/p definitions, as it's not possible
+ to determine the revision from the device ID.
+ - Same for the Tseng ET4000.
+ The id's said I had a rev C, when it was actually a rev B.
+ - Removed some unnecessary strings
+ `PCI'. Unless it's an AGP card, it should be obvious that it's PCI.
+ `Ethernet card'. This should be determined from config space.
+ - Lots of other new vendors & devices
+ - Other additional small cleanups.
+
+Mon Jun 21 22:11:10 1999 Martin Mares <mj@albireo.ucw.cz>
+
+ * lib/names.c (pci_lookup_name): When device ID lookup
+ fails, return hexadecimal device ID instead of vendor ID.
+ Well spotted by Axel Bodemann <Axel.Bodemann@ruhr-uni-bochum.de>,
+ I've just extended the fix to subsystem ID's.
+
Thu Jun 17 19:48:21 1999 Martin Mares <mj@albireo.ucw.cz>
* lspci.c (show_agp): Bug fix from Petr Vandrovec
(show_htype2): Stub routine.
(scan_config): Write sensible error message if the kernel denies reading of
upper part of the PCI config space.
-
projects / pciutils.git / blobdiff