2009-07-04 Martin Mares <mj@ucw.cz>
* Released as 3.1.3.
* Updated pci.ids to the current snapshot of the database.
* The VPD parser now reports unknown and vendor-defined items
properly. It also stops on any item in unknown format, avoiding long
output on bogus VPD data. Thanks to Ben Hutchings and Matthew Wilcox.
* The MSI-X table size now matches the spec. Thanks to Michael S.
Tsirkin.
* The Power Management capability now includes the soft reset bit.
Thanks to Yu Zhao.
* Decoding of the Advanced Features capability has been added.
Thanks to Yu Zhao.
* The whole package compiles on GNU/kFreeBSD again.
The following patches have been contributed by Matthew Wilcox:
* The procfs back-end is able to cope with /proc/bus/pci
containing names with domains, which occur on sparc64 and
possibly other architectures due to a kernel bug.
* The sysfs back-end no longer complains when a slot address
is missing, which happens with old versions of Linux fakephp.
* The Device Serial Number capability is printed in the right
byte order.
* The MSI and MSI-X capabilities are printed in a prettier way.
* The tree output mode (`lspci -t') shows domain numbers only
at the root, which makes the output more compact.
* Updated documentation on the bus mapping mode (`lspci -M').
2009-02-01 Martin Mares <mj@ucw.cz>
* Released as 3.1.2.
* Fixed another silly bug in the command-line parser of setpci.
2009-01-30 Martin Mares <mj@ucw.cz>
* Released as 3.1.1.
* Updated pci.ids to the current snapshot of the database.
* The configure script now sets LC_ALL to avoid being fooled by
locale-dependent behavior of `tr'.
* The command-line parser of setpci did sometimes segfault on invalid
input. Thanks to Andreas Wiese for a fix.
2009-01-18 Martin Mares <mj@ucw.cz>
* Released as 3.1.0.
* Updated pci.ids to the current snapshot of the database.
* The Cygwin backend now works on Windows Vista. Thanks to Jonathan
Kotta.
* Fixed a bug in decoding of the SR-IOV capability. Patch by Yu Zhao.
* Details of some PCIe capabilities are displayed only with -vv.
* When a BAR is reported by the OS, but not by the device (i.e.,
it is marked as [virtual] in lspci), the [disabled] flag is
suppressed, because it does not make sense in such cases.
Patch by Yu Zhao.
2008-12-13 Martin Mares <mj@ucw.cz>
* The source code of lspci has been split to multiple files, hopefully
making it easier to maintain.
* The library and lspci now know about physical slot names. So far,
they are provided by the sysfs back-end only. Thanks go to Alex Chiang.
* When a device has the VPD (Vital Product Data) capability and the
VPD data are supplied by the OS, they are decoded and printed in the
verbose mode. This currently works only on Linux with the sysfs
back-end. Thanks to Ben Hutchings of Solarflare for the patch.
* `setpci --version' now works properly.
* `setpci --dumpregs' prints a table of all known names of
registers and capabilities. This replaces the table of registers
in the setpci man page.
* The dry-run mode of setpci gives better feedback.
* The setpci utility is now able to address registers stored in PCI
capabilities (actually it allows a more general form of relative
addressing).
* The library has gained functions for working with PCI capabilities.
* Address Translation Services capability is now decoded. Patch by
Yu Zhao.
2008-11-09 Martin Mares <mj@ucw.cz>
* Released as 3.0.3.
* `lspci -k' now displays the subsystem ID, too. This makes `-k'
show everything needed to identify the device and the available
drivers, which was called for by many users.
* Fixed spelling of MSI. Patch by Matthew Wilcox.
* Better support for cross-compilation. Thanks to Alon Bar-Lev
for the patch.
* Fixed printing of the AER capability. Patch by Max Asbock.
* HT 1.02 capabilities are decoded as HT 1.03. Suggested by
Carl-Daniel Hailfinger.
* Fixed Cygwin build. Thanks to Steve Elliott for reporting the bug.
* Updated pci.ids to the current snapshot of the database.
This changes the buildlink3.mk files to use an include guard for the
recursive include. The use of BUILDLINK_DEPTH, BUILDLINK_DEPENDS,
BUILDLINK_PACKAGES and BUILDLINK_ORDER is handled by a single new
variable BUILDLINK_TREE. Each buildlink3.mk file adds a pair of
enter/exit marker, which can be used to reconstruct the tree and
to determine first level includes. Avoiding := for large variables
(BUILDLINK_ORDER) speeds up parse time as += has linear complexity.
The include guard reduces system time by avoiding reading files over and
over again. For complex packages this reduces both %user and %sys time to
half of the former time.
and add a new helper target and script, "show-buildlink3", that outputs
a listing of the buildlink3.mk files included as well as the depth at
which they are included.
For example, "make show-buildlink3" in fonts/Xft2 displays:
zlib
fontconfig
iconv
zlib
freetype2
expat
freetype2
Xrender
renderproto
RECOMMENDED is removed. It becomes ABI_DEPENDS.
BUILDLINK_RECOMMENDED.foo becomes BUILDLINK_ABI_DEPENDS.foo.
BUILDLINK_DEPENDS.foo becomes BUILDLINK_API_DEPENDS.foo.
BUILDLINK_DEPENDS does not change.
IGNORE_RECOMMENDED (which defaulted to "no") becomes USE_ABI_DEPENDS
which defaults to "yes".
Added to obsolete.mk checking for IGNORE_RECOMMENDED.
I did not manually go through and fix any aesthetic tab/spacing issues.
I have tested the above patch on DragonFly building and packaging
subversion and pkglint and their many dependencies.
I have also tested USE_ABI_DEPENDS=no on my NetBSD workstation (where I
have used IGNORE_RECOMMENDED for a long time). I have been an active user
of IGNORE_RECOMMENDED since it was available.
As suggested, I removed the documentation sentences suggesting bumping for
"security" issues.
As discussed on tech-pkg.
I will commit to revbump, pkglint, pkg_install, createbuildlink separately.
Note that if you use wip, it will fail! I will commit to pkgsrc-wip
later (within day).
