From: Knut Omang <knut.omang@oracle.com>
To: marcel@redhat.com, qemu-devel@nongnu.org
Cc: Eduardo Habkost <ehabkost@redhat.com>,
"Michael S. Tsirkin" <mst@redhat.com>,
"Richard W.M. Jones" <rjones@redhat.com>,
Alex Williamson <alex.williamson@redhat.com>,
"Gonglei (Arei)" <arei.gonglei@huawei.com>,
Jan Kiszka <jan.kiszka@web.de>,
Paolo Bonzini <pbonzini@redhat.com>,
Dotan Barak <dotanba@gmail.com>,
Richard Henderson <rth@twiddle.net>
Subject: Re: [Qemu-devel] [PATCH v5 2/4] pcie: Add support for Single Root I/O Virtualization (SR/IOV)
Date: Sun, 18 Oct 2015 17:57:04 +0200 [thread overview]
Message-ID: <1445183824.7681.146.camel@oracle.com> (raw)
In-Reply-To: <56237C62.3050909@gmail.com>
On Sun, 2015-10-18 at 14:02 +0300, Marcel Apfelbaum wrote:
> > +#define SRIOV_ID(dev) \
> > + (dev)->name, PCI_SLOT((dev)->devfn), PCI_FUNC((dev)->devfn)
>
> This is a little "hacky" but it is used only for tracing and inside a
> C file.
> I am OK with it.
For the sake of readability of the trace calls and to avoid the trace
lines from disturbing the code overview too much - I would have had to
split the line of each trace call.
> > +++ b/include/hw/pci/pcie_sriov.h
> > @@ -0,0 +1,58 @@
> > +/*
> > + * pcie_sriov.h:
> > + *
> > + * Implementation of SR/IOV emulation support.
> > + *
> > + * Copyright (c) 2015 Knut Omang <knut.omang@oracle.com>
> > + *
> > + * This work is licensed under the terms of the GNU GPL, version 2
> > or later.
> > + * See the COPYING file in the top-level directory.
> > + *
> > + */
> > +
> > +#ifndef QEMU_PCIE_SRIOV_H
> > +#define QEMU_PCIE_SRIOV_H
> > +
> > +struct PCIESriovPF {
> > + uint16_t num_vfs; /* Number of virtual functions
> > created */
> > + uint8_t vf_bar_type[PCI_NUM_REGIONS]; /* Store type for each
> > VF bar */
> > + const char *vfname; /* Reference to the device type
> > used for the VFs */
> > + PCIDevice **vf; /* Pointer to an array of num_vfs
> > VF devices */
> > +};
> > +
> > +struct PCIESriovVF {
> > + PCIDevice *pf; /* Pointer back to owner physical
> > function */
> > +};
>
>
> Regarding naming conventions, maybe you it should be PCIESRIOVPF o Ir
> PCIESRIOVPf,
> but this would not justify another version (in my opinion)
Yes, I did think about that, but didn't like the readability of either
of these, still wanted to try to follow conventions - this was the
tradeoff I ended at. SR/IOV is kind of complicated compared to e.g. AER
in that the '/' has to go anyway :-D
> > +
> > +/* Optionally add supported page sizes to the mask of supported
> > page sizes
> > + * Page size values are interpreted as opt_sup_pgsize << 12.
> > + */
> > +void pcie_sriov_pf_add_sup_pgsize(PCIDevice *dev, uint16_t
> > opt_sup_pgsize);
> > +
> > +/* SR/IOV capability config write handler */
> > +void pcie_sriov_config_write(PCIDevice *dev, uint32_t address,
> > + uint32_t val, int len);
> > +
> > +/* Reset SR/IOV VF Enable bit to unregister all VFs */
> > +void pcie_sriov_pf_disable_vfs(PCIDevice *dev);
> > +
> > +#endif /* QEMU_PCIE_SRIOV_H */
> > diff --git a/include/qemu/typedefs.h b/include/qemu/typedefs.h
> > index ee1ce1d..0b89316 100644
> > --- a/include/qemu/typedefs.h
> > +++ b/include/qemu/typedefs.h
> > @@ -54,6 +54,8 @@ typedef struct PCIDevice PCIDevice;
> > typedef struct PCIEAERErr PCIEAERErr;
> > typedef struct PCIEAERLog PCIEAERLog;
> > typedef struct PCIEAERMsg PCIEAERMsg;
> > +typedef struct PCIESriovPF PCIESriovPF;
> > +typedef struct PCIESriovVF PCIESriovVF;
> > typedef struct PCIEPort PCIEPort;
> > typedef struct PCIESlot PCIESlot;
> > typedef struct PCIExpressDevice PCIExpressDevice;
> > diff --git a/trace-events b/trace-events
> > index a0ddc6b..fd51c5f 100644
> > --- a/trace-events
> > +++ b/trace-events
> > @@ -1531,6 +1531,11 @@ xen_pv_mmio_write(uint64_t addr) "WARNING:
> > write to Xen PV Device MMIO space (ad
> > pci_cfg_read(const char *dev, unsigned devid, unsigned fnid,
> > unsigned offs, unsigned val) "%s %02u:%u @0x%x -> 0x%x"
> > pci_cfg_write(const char *dev, unsigned devid, unsigned fnid,
> > unsigned offs, unsigned val) "%s %02u:%u @0x%x <- 0x%x"
> >
> > +# hw/pci/pcie_sriov.c
> > +sriov_register_vfs(const char *name, int slot, int function, int
> > num_vfs) "%s %02x:%x: creating %d vf devs"
> > +sriov_unregister_vfs(const char *name, int slot, int function, int
> > num_vfs) "%s %02x:%x: Unregistering %d vf devs"
> > +sriov_config_write(const char *name, int slot, int fun, uint32_t
> > offset, uint32_t val, uint32_t len) "%s %02x:%x: sriov offset 0x%x
> > val 0x%x len %d"
> > +
> > # hw/vfio/pci.c
> > vfio_intx_interrupt(const char *name, char line) " (%s) Pin %c"
> > vfio_intx_eoi(const char *name) " (%s) EOI"
> >
>
> Thanks for all the work, besides the questions above, from my point
> of view:
>
>
> Reviewed-by: Marcel Apfelbaum <marcel@redhat.com>
Thanks,
Knut
> Thanks,
> Marcel
next prev parent reply other threads:[~2015-10-18 15:57 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-10-14 15:51 [Qemu-devel] [PATCH v5 0/4] pcie: Add support for Single Root I/O Virtualization Knut Omang
2015-10-14 15:51 ` [Qemu-devel] [PATCH v5 1/4] pci: Make use of the devfn property when registering new devices Knut Omang
2015-10-14 15:51 ` [Qemu-devel] [PATCH v5 2/4] pcie: Add support for Single Root I/O Virtualization (SR/IOV) Knut Omang
2015-10-18 11:02 ` Marcel Apfelbaum
2015-10-18 12:26 ` Michael S. Tsirkin
2015-10-18 15:00 ` Knut Omang
2015-10-18 15:57 ` Knut Omang [this message]
2015-10-19 9:00 ` Paolo Bonzini
2015-10-14 15:51 ` [Qemu-devel] [PATCH v5 3/4] pcie: Add some SR/IOV API documentation in docs/pcie_sriov.txt Knut Omang
2015-10-16 8:36 ` Michael S. Tsirkin
2015-10-16 9:56 ` Knut Omang
2015-10-16 10:54 ` Michael S. Tsirkin
2015-10-16 11:25 ` Knut Omang
2015-10-16 11:34 ` Michael S. Tsirkin
2015-10-16 11:39 ` Paolo Bonzini
2015-10-16 11:27 ` Marcel Apfelbaum
2015-10-16 11:32 ` Richard W.M. Jones
2015-10-16 11:39 ` Michael S. Tsirkin
2015-10-18 11:03 ` Marcel Apfelbaum
2015-10-14 15:51 ` [Qemu-devel] [PATCH v5 4/4] pcie: A few minor fixes (type+code simplify) Knut Omang
2015-10-18 11:03 ` Marcel Apfelbaum
2015-10-18 11:02 ` [Qemu-devel] [PATCH v5 0/4] pcie: Add support for Single Root I/O Virtualization Marcel Apfelbaum
2015-10-18 14:39 ` Knut Omang
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=1445183824.7681.146.camel@oracle.com \
--to=knut.omang@oracle.com \
--cc=alex.williamson@redhat.com \
--cc=arei.gonglei@huawei.com \
--cc=dotanba@gmail.com \
--cc=ehabkost@redhat.com \
--cc=jan.kiszka@web.de \
--cc=marcel@redhat.com \
--cc=mst@redhat.com \
--cc=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=rjones@redhat.com \
--cc=rth@twiddle.net \
/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).