From mboxrd@z Thu Jan  1 00:00:00 1970
From: Peter =?utf-8?q?M=C3=BCller?= <peter.mueller@ipfire.org>
To: development@lists.ipfire.org
Subject: Re: [PATCH] pciutils: Update to version 3.7.0
Date: Sun, 06 Feb 2022 10:47:09 +0000
Message-ID: <97178fdf-969c-c961-bb18-b9abc41de8ef@ipfire.org>
In-Reply-To: <94b05e32-db46-5be9-a555-441307bb81b7@ipfire.org>
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="===============4403135473799205846=="
List-Id: <development.lists.ipfire.org>

--===============4403135473799205846==
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable

Hello Adolf,

thanks for your reply.

> Hi Peter,
>=20
> One difference is that I changed from the .gz file to the .xz file

*sigh* I overlooked that one due to lack of coffee. Sorry.

> However I had also downloaded from the
> https://mirrors.edge.kernel.org/pub/software/utils/pciutils/ site
> This has both .gz and .xz files.

In generally, xz provides a better compression rate than gz, so it is better =
to use
such a file if available. However, pciutils will not be the straw that breaks=
 the
camel's neck here. :-)

> I just now checked the md5sum from the .gz from this location and it is sti=
ll different from the one you quoted.
>=20
> You downloaded from the
> https://mj.ucw.cz/download/linux/pci/ site. This only has a .gz file and do=
ing an md5sum on that file I get the same as you.

Yes, I downloaded it from there, since this the project's homepage.

> So there is a difference in the md5sum for the supposedly same file stored =
on the mj.ucw.cz location compared to the www.kernel.org location.

Again... :-/

> The reason I used the www.kernel.org location is that they have a whole ran=
ge of packages on their location covering various core stuff and seem to be a=
 common location for that sort of stuff. Many packages only get stored on www=
.kernel.org (or its mirrors). I think this is the first that I have found hav=
ing a local copy as well.

Oh, that's good to know. I sometimes had trouble finding a download source fo=
r the
tarball. Usually, I searched on https://www.debian.org/distrib/packages and t=
ried to
make my way from there...

> It will be interesting to hear what you find out for the reason for the dif=
ference.

Indeed. I will have a look at this later.

Thanks, and best regards,
Peter M=C3=BCller

>=20
> Just checked and Arch Linux used to download the source from the www.kernel=
.org location but that source line is now commented out and the mj.ucw.cz loc=
ation is now being used in their PKGBUILD.
>=20
> Regards,
> Adolf.
>=20
>=20
> On 04/02/2022 13:05, Peter M=C3=BCller wrote:
>> Hello Adolf,
>>
>> I regret to inform you that I worked on the same thing.
>>
>> Comparing both patches (see https://git.ipfire.org/?p=3Dipfire-2.x.git;a=
=3Dcommit;h=3D34a98832ee5ec33ef75f4732dfed003e9b76185c
>> for mine), the MD5 checksum differs: Yours is e6e20482b4f25c5186e6a753c5ed=
c361, while
>> mine is 51554c538b5a57b61123326e14ea28a1.
>>
>> I will investigate into this and report back later.
>>
>> Thanks, and best regards,
>> Peter M=C3=BCller
>>
>>
>>> - Update from 3.5.6 (2017) to 3.7.0 (2020)
>>> - Update of rootfile
>>> - Changelog
>>> =C2=A0=C2=A0=C2=A0=C2=A0 2020-05-31=C2=A0 Martin Mares <mj(a)ucw.cz>
>>> =C2=A0=C2=A0=C2=A0=C2=A0 * Released as 3.7.0.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* Added or improved the following capabilities: D=
esignated Vendor-Specific,
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 Compute eXpress Link, Resizable BARs, VF R=
esizable BARs, Link
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 Capabilities 2, Link Status 2.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* On Linux, lspci can show IOMMU groups.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* setpci can be asked to skip bus scan and operat=
e on a device
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 completely specified by its domain/bus/dev=
/func address. This
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 involved major internal cleanup.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* The above feature of setpci uses the pci_get_de=
v() function,
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 which obtains a struct pci_dev without doi=
ng a bus scan. This was
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 always possible, but apparently little use=
d, because back-ends
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 frequently choked when operating on such d=
evices. Fixed a lot
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 of minor bugs related to this.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* Also, back-ends which do not support domains no=
w correctly fail when
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 trying to access devices outside domain 0.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* Semantics of pci_fill_info() and pci_dev->known=
_fields was underspecified,
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 which lead to inconsistencies between back=
-ends. Improved documentation
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 to give a more precise definition and upda=
ted all back-ends to conform
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 to it. Most importantly, pci_dev->known_fi=
elds shows all fields requested
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 over the lifetime of the pci_dev, but neve=
r those which are not supported
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 by the back-end.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* As usually, updated pci.ids to the current snap=
shot of the database.
>>> =C2=A0=C2=A0=C2=A0=C2=A0 2020-01-25=C2=A0 Martin Mares <mj(a)ucw.cz>
>>> =C2=A0=C2=A0=C2=A0=C2=A0 * Released as 3.6.4.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* A new back-end for the GNU Hurd was contributed=
 by Joan Lled=C3=B3.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* When printing VPD item identifiers, non-ASCII c=
haracters are escaped.
>>> =C2=A0=C2=A0=C2=A0=C2=A0 2020-01-22=C2=A0 Martin Mares <mj(a)ucw.cz>
>>> =C2=A0=C2=A0=C2=A0=C2=A0 * Released as 3.6.3.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* `lspci -t' (tree mode) can be combined with `-s=
' to show a sub-tree.
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 We also fixed potential buffer overflows i=
n the tree dumper.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* Cleaned messy code for dumping of I/O, memory, =
and ROM regions.
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 This helped fixing a bug, which caused som=
e 64-bit regions to be
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 reported as virtual. All flags are now pri=
nted after the address
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 (previously, "[virtual]" and "[enhanced]" =
were before it for no good
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 reason).
>>> =C2=A0=C2=A0=C2=A0=C2=A0* Added pci_find_cap_nr() to the library, which h=
andles capabilities
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 which occur multiple times in a single dev=
ice.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* Minor improvements in printing of PCIe capabili=
ties.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* We now decode the Multicast and Secondary PCI E=
xpress extended
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 capabilities.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* The list of capability names available to setpc=
i was updated.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* Minor bugs were fixed in FreeBSD and Solaris po=
rts.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* We now prefer HTTPS URLs in all documentation
>>> =C2=A0=C2=A0=C2=A0=C2=A0* The pci.ids file has a man page.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* As usually, updated pci.ids to the current snap=
shot of the database.
>>> =C2=A0=C2=A0=C2=A0=C2=A0 2018-08-12=C2=A0 Martin Mares <mj(a)ucw.cz>
>>> =C2=A0=C2=A0=C2=A0=C2=A0 * Released as 3.6.2.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* Added "-P" and "-PP" switches to lspci, which d=
isplay the path
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 through bridges to each device.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* Fixed a couple of bugs in computation of bus to=
pology. It was
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 previously used only for the tree display,=
 but we re-use it for
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 computing the paths.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* As usually, updated pci.ids to the current snap=
shot of the database.
>>> =C2=A0=C2=A0=C2=A0=C2=A0 2018-07-12=C2=A0 Martin Mares <mj(a)ucw.cz>
>>> =C2=A0=C2=A0=C2=A0=C2=A0 * Released as 3.6.1.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* Fixed compilation issues on Linux systems with =
MUSL libc.
>>> =C2=A0=C2=A0=C2=A0=C2=A0 2018-06-30=C2=A0 Martin Mares <mj(a)ucw.cz>
>>> =C2=A0=C2=A0=C2=A0=C2=A0 * Released as 3.6.0.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* BARs reported by the OS, but not set on the dev=
ice itself are
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 reliably marked with "[virtual]".
>>> =C2=A0=C2=A0=C2=A0=C2=A0* Library: Introduced a generic mechanism of stri=
ng properties. This
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 avoids lots of special cases and makes ABI=
 compatibility easier.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* On Linux systems with OpenFirmware, report corr=
esponding device tree nodes
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 as device properties.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* VPD decoder knows several non-standard extensio=
ns.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* When PCIe link speed is less than the maximum s=
upported by the
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 device, it is explicitly marked as "downgr=
aded".
>>> =C2=A0=C2=A0=C2=A0=C2=A0* Several new capabilities are not decoded yet, b=
ut at least their
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 names are printed.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* The Null capability is easily decoded.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* Formatting of several capabilities was cleaned =
up.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* The VGA16 bit in the bridge control register is=
 now supported.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* Added a port to SylixOS. Thanks to YuJian Gong =
for contribution.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* Added a port to DOS/DJGPP. Thanks to Rudolf Mar=
ek for contribution.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* The order in which back-ends are probed was dec=
oupled from the
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 internal back-end IDs. This helps, because=
 new back-ends must have
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 their ID allocated at the end to keep the =
ABI, but they might need
>>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 to be probed earlier.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* The fbsd-device back-end should work again.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* Fixed a couple of bugs. Most notably, DeviceNam=
e was not printed.
>>> =C2=A0=C2=A0=C2=A0=C2=A0* As usually, updated pci.ids to the current snap=
shot of the database.
>>>
>>> Signed-off-by: Adolf Belka <adolf.belka(a)ipfire.org>
>>> ---
>>> =C2=A0 config/rootfiles/common/pciutils | 2 +-
>>> =C2=A0 lfs/pciutils=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 | 6 +++---
>>> =C2=A0 2 files changed, 4 insertions(+), 4 deletions(-)
>>>
>>> diff --git a/config/rootfiles/common/pciutils b/config/rootfiles/common/p=
ciutils
>>> index da35ef6a1..17a3ede66 100644
>>> --- a/config/rootfiles/common/pciutils
>>> +++ b/config/rootfiles/common/pciutils
>>> @@ -1,7 +1,7 @@
>>> =C2=A0 bin/lspci
>>> =C2=A0 bin/setpci
>>> =C2=A0 lib/libpci.so.3
>>> -lib/libpci.so.3.5.6
>>> +lib/libpci.so.3.7.0
>>> =C2=A0 sbin/update-pciids
>>> =C2=A0 #usr/include/pci
>>> =C2=A0 #usr/include/pci/config.h
>>> diff --git a/lfs/pciutils b/lfs/pciutils
>>> index 2068c5a0c..2bd5d3c5d 100644
>>> --- a/lfs/pciutils
>>> +++ b/lfs/pciutils
>>> @@ -24,10 +24,10 @@
>>> =C2=A0 =C2=A0 include Config
>>> =C2=A0 -VER=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 =3D 3.5.6
>>> +VER=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 =3D 3.7.0
>>> =C2=A0 =C2=A0 THISAPP=C2=A0=C2=A0=C2=A0 =3D pciutils-$(VER)
>>> -DL_FILE=C2=A0=C2=A0=C2=A0 =3D $(THISAPP).tar.gz
>>> +DL_FILE=C2=A0=C2=A0=C2=A0 =3D $(THISAPP).tar.xz
>>> =C2=A0 DL_FROM=C2=A0=C2=A0=C2=A0 =3D $(URL_IPFIRE)
>>> =C2=A0 DIR_APP=C2=A0=C2=A0=C2=A0 =3D $(DIR_SRC)/$(THISAPP)
>>> =C2=A0 TARGET=C2=A0=C2=A0=C2=A0=C2=A0 =3D $(DIR_INFO)/$(THISAPP)
>>> @@ -40,7 +40,7 @@ objects =3D $(DL_FILE)
>>> =C2=A0 =C2=A0 $(DL_FILE) =3D $(DL_FROM)/$(DL_FILE)
>>> =C2=A0 -$(DL_FILE)_MD5 =3D 63b347d2bb5719b7769c03a1bb72a5b7
>>> +$(DL_FILE)_MD5 =3D e6e20482b4f25c5186e6a753c5edc361
>>> =C2=A0 =C2=A0 install : $(TARGET)
>>> =C2=A0=20
>=20

--===============4403135473799205846==--