From: Luca Vizzarro <Luca.Vizzarro@arm.com>
To: jspewock@iol.unh.edu, npratte@iol.unh.edu,
juraj.linkes@pantheon.tech, yoan.picchi@foss.arm.com,
thomas@monjalon.net, Honnappa.Nagarahalli@arm.com,
probb@iol.unh.edu, wathsala.vithanage@arm.com,
paul.szczepanek@arm.com, alex.chapman@arm.com
Cc: dev@dpdk.org
Subject: Re: [PATCH v4 4/5] dts: add OS abstractions for creating virtual functions
Date: Thu, 14 Nov 2024 17:29:12 +0000 [thread overview]
Message-ID: <54f1e692-e788-40ce-85e4-f4b9fa92cd72@arm.com> (raw)
In-Reply-To: <20240923184235.22582-5-jspewock@iol.unh.edu>
Usual problem with the commit subject.
On 23/09/2024 19:42, jspewock@iol.unh.edu wrote:
> + def set_num_virtual_functions(self, num: int, pf_port: Port) -> bool:
I'd call this set_num_of_virtual_functions, set_virtual_functions_num,
or set_number_of_vfs. I think I prefer the last one.
> + """Overrides :meth:`~.os_session.OSSession.set_num_virtual_functions`."""
> + sys_bus_path = f"/sys/bus/pci/devices/{pf_port.pci}/sriov_numvfs".replace(":", "\\:")
> + curr_num_vfs = int(self.send_command(f"cat {sys_bus_path}").stdout)
> + if num > 0 and curr_num_vfs >= num:
> + self._logger.info(
> + f"{curr_num_vfs} VFs already configured on port {pf_port.identifier.pci} on node "
> + f"{pf_port.identifier.node}."
> + )
> + return False
> + elif num > 0 and curr_num_vfs > 0:
> + self._logger.error(
> + f"Not enough VFs configured on port {pf_port.identifier.pci} on node "
> + f"{pf_port.identifier.node}. Need {num} but only {curr_num_vfs} are configured. "
> + "DTS is unable to modify number of preexisting VFs."
> + )
> + raise InternalError("Failed to create VFs on port.")
This is not an InternalError. That exception represents a bug internal
to DTS, it's not related to the node not satisfying requirements. I am
guessing this requires a new kind of error.
> + self.send_command(f"echo {num} > {sys_bus_path}", privileged=True, verify=True)
> + return True
> +
> + def get_pci_addr_of_vfs(self, pf_port: Port) -> list[str]:
> + """Overrides :meth:`~.os_session.OSSession.get_pci_addr_of_vfs`."""
> + sys_bus_path = f"/sys/bus/pci/devices/{pf_port.pci}".replace(":", "\\:")
> + curr_num_vfs = int(self.send_command(f"cat {sys_bus_path}/sriov_numvfs").stdout)
> + if curr_num_vfs > 0:
> + pci_addrs = self.send_command(
> + 'awk -F "PCI_SLOT_NAME=" "/PCI_SLOT_NAME=/ {print \\$2}" '
> + + f"{sys_bus_path}/virtfn*/uevent",
I like the use of awk. Using a TextWrapper may be overkill when we have
easier options like this one.
> + def set_num_virtual_functions(self, num: int, pf_port: Port) -> bool:
> + """Update the number of virtual functions (VFs) on a port.
Like Juraj mentioned, this docstring should say that it can be used for
creating and removing VFs.
next prev parent reply other threads:[~2024-11-14 17:29 UTC|newest]
Thread overview: 42+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-08-21 19:15 [RFC PATCH v1 0/5] dts: add VFs to the framework jspewock
2024-08-21 19:15 ` [RFC PATCH v1 1/5] dts: allow binding only a single port to a different driver jspewock
2024-08-21 19:15 ` [RFC PATCH v1 2/5] dts: parameterize what ports the TG sends packets to jspewock
2024-08-21 19:15 ` [RFC PATCH v1 3/5] dts: add class for virtual functions jspewock
2024-08-21 19:15 ` [RFC PATCH v1 4/5] dts: add OS abstractions for creating " jspewock
2024-08-21 19:15 ` [RFC PATCH v1 5/5] dts: add functions for managing VFs to Node jspewock
2024-08-21 19:21 ` [RFC PATCH v2 0/5] dts: add VFs to the framework jspewock
2024-08-21 19:21 ` [RFC PATCH v2 1/5] dts: allow binding only a single port to a different driver jspewock
2024-08-21 19:21 ` [RFC PATCH v2 2/5] dts: parameterize what ports the TG sends packets to jspewock
2024-08-21 19:21 ` [RFC PATCH v2 3/5] dts: add class for virtual functions jspewock
2024-08-21 19:21 ` [RFC PATCH v2 4/5] dts: add OS abstractions for creating " jspewock
2024-08-21 19:21 ` [RFC PATCH v2 5/5] dts: add functions for managing VFs to Node jspewock
2024-08-21 19:38 ` [RFC PATCH v2 0/5] dts: add VFs to the framework jspewock
2024-08-21 19:38 ` [RFC PATCH v2 1/5] dts: allow binding only a single port to a different driver jspewock
2024-08-21 19:38 ` [RFC PATCH v2 2/5] dts: parameterize what ports the TG sends packets to jspewock
2024-08-21 19:38 ` [RFC PATCH v2 3/5] dts: add class for virtual functions jspewock
2024-08-21 19:38 ` [RFC PATCH v2 4/5] dts: add OS abstractions for creating " jspewock
2024-08-21 19:38 ` [RFC PATCH v2 5/5] dts: add functions for managing VFs to Node jspewock
2024-08-21 19:44 ` [RFC PATCH v2 0/5] dts: add VFs to the framework Jeremy Spewock
2024-08-21 21:30 ` [RFC PATCH v3 " jspewock
2024-08-21 21:30 ` [RFC PATCH v3 1/5] dts: allow binding only a single port to a different driver jspewock
2024-08-21 21:30 ` [RFC PATCH v3 2/5] dts: parameterize what ports the TG sends packets to jspewock
2024-08-21 21:30 ` [RFC PATCH v3 3/5] dts: add class for virtual functions jspewock
2024-08-21 21:30 ` [RFC PATCH v3 4/5] dts: add OS abstractions for creating " jspewock
2024-08-21 21:30 ` [RFC PATCH v3 5/5] dts: add functions for managing VFs to Node jspewock
2024-09-23 18:42 ` [PATCH v4 0/5] dts: add VFs to the framework jspewock
2024-09-23 18:42 ` [PATCH v4 1/5] dts: allow binding only a single port to a different driver jspewock
2024-09-25 8:45 ` Juraj Linkeš
2024-11-14 16:45 ` Luca Vizzarro
2024-09-23 18:42 ` [PATCH v4 2/5] dts: parameterize what ports the TG sends packets to jspewock
2024-09-25 10:58 ` Juraj Linkeš
2024-11-14 17:01 ` Luca Vizzarro
2024-09-23 18:42 ` [PATCH v4 3/5] dts: add class for virtual functions jspewock
2024-09-25 11:28 ` Juraj Linkeš
2024-11-14 17:10 ` Luca Vizzarro
2024-09-23 18:42 ` [PATCH v4 4/5] dts: add OS abstractions for creating " jspewock
2024-09-25 12:05 ` Juraj Linkeš
2024-11-14 17:29 ` Luca Vizzarro [this message]
2024-09-23 18:42 ` [PATCH v4 5/5] dts: add functions for managing VFs to Node jspewock
2024-09-25 13:29 ` Juraj Linkeš
2024-11-14 17:36 ` Luca Vizzarro
2024-09-25 8:24 ` [PATCH v4 0/5] dts: add VFs to the framework Juraj Linkeš
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=54f1e692-e788-40ce-85e4-f4b9fa92cd72@arm.com \
--to=luca.vizzarro@arm.com \
--cc=Honnappa.Nagarahalli@arm.com \
--cc=alex.chapman@arm.com \
--cc=dev@dpdk.org \
--cc=jspewock@iol.unh.edu \
--cc=juraj.linkes@pantheon.tech \
--cc=npratte@iol.unh.edu \
--cc=paul.szczepanek@arm.com \
--cc=probb@iol.unh.edu \
--cc=thomas@monjalon.net \
--cc=wathsala.vithanage@arm.com \
--cc=yoan.picchi@foss.arm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).