From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:52680) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1dRxph-0008Gk-8y for qemu-devel@nongnu.org; Mon, 03 Jul 2017 05:39:54 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1dRxpc-00071v-AT for qemu-devel@nongnu.org; Mon, 03 Jul 2017 05:39:53 -0400 Received: from mx1.redhat.com ([209.132.183.28]:46262) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1dRxpc-000719-16 for qemu-devel@nongnu.org; Mon, 03 Jul 2017 05:39:48 -0400 Date: Mon, 3 Jul 2017 11:39:40 +0200 From: Igor Mammedov Message-ID: <20170703113940.0e0415a2@nial.brq.redhat.com> In-Reply-To: <1498745240-30658-6-git-send-email-mark.cave-ayland@ilande.co.uk> References: <1498745240-30658-1-git-send-email-mark.cave-ayland@ilande.co.uk> <1498745240-30658-6-git-send-email-mark.cave-ayland@ilande.co.uk> MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCHv7 5/6] fw_cfg: move qdev_init_nofail() from fw_cfg_init1() to callers List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Mark Cave-Ayland Cc: qemu-devel@nongnu.org, lersek@redhat.com, somlo@cmu.edu, ehabkost@redhat.com, mst@redhat.com, pbonzini@redhat.com, rjones@redhat.com, peter.maydell@linaro.org On Thu, 29 Jun 2017 15:07:19 +0100 Mark Cave-Ayland wrote: > When looking to instantiate a TYPE_FW_CFG_MEM or TYPE_FW_CFG_IO device to be > able to wire it up differently, it is much more convenient for the caller to > instantiate the device and have the fw_cfg default files already preloaded > during realize. > > Move fw_cfg_init1() to the end of both the fw_cfg_mem_realize() and > fw_cfg_io_realize() functions so it no longer needs to be called manually > when instantiating the device, and also rename it to fw_cfg_common_realize() > which better describes its new purpose. > > Since it is now the responsibility of the machine to wire up the fw_cfg device > it is necessary to introduce a object_property_add_child() call into > fw_cfg_init_io() and fw_cfg_init_mem() to link the fw_cfg device to the root > machine object as before. > > Finally we can now convert the asserts() preventing multiple fw_cfg devices > and unparented fw_cfg devices being instantiated and replace them with proper > error reporting at realize time. This allows us to remove FW_CFG_NAME and > FW_CFG_PATH since they are no longer required. > > Signed-off-by: Mark Cave-Ayland > --- > hw/nvram/fw_cfg.c | 41 +++++++++++++++++++++++++++++------------ > 1 file changed, 29 insertions(+), 12 deletions(-) > > diff --git a/hw/nvram/fw_cfg.c b/hw/nvram/fw_cfg.c > index 2291121..31029ac 100644 > --- a/hw/nvram/fw_cfg.c > +++ b/hw/nvram/fw_cfg.c > @@ -37,9 +37,6 @@ > > #define FW_CFG_FILE_SLOTS_DFLT 0x20 > > -#define FW_CFG_NAME "fw_cfg" > -#define FW_CFG_PATH "/machine/" FW_CFG_NAME > - > #define TYPE_FW_CFG "fw_cfg" > #define TYPE_FW_CFG_IO "fw_cfg_io" > #define TYPE_FW_CFG_MEM "fw_cfg_mem" > @@ -920,19 +917,22 @@ static int fw_cfg_unattached_at_realize(void) > } > > > -static void fw_cfg_init1(DeviceState *dev) > +static void fw_cfg_common_realize(DeviceState *dev, Error **errp) > { > FWCfgState *s = FW_CFG(dev); > MachineState *machine = MACHINE(qdev_get_machine()); > uint32_t version = FW_CFG_VERSION; > > - assert(!object_resolve_path(FW_CFG_PATH, NULL)); > - > - object_property_add_child(OBJECT(machine), FW_CFG_NAME, OBJECT(s), NULL); > - > - qdev_init_nofail(dev); > + if (!fw_cfg_find()) { maybe add comment that here, that fw_cfg_find() will return NULL if more than 1 device is found. > + error_setg(errp, "at most one %s device is permitted", TYPE_FW_CFG); s/TYPE_FW_CFG/object_get_typename()/ so it would print leaf type name > + return; > + } > > - assert(!fw_cfg_unattached_at_realize()); > + if (fw_cfg_unattached_at_realize()) { as I pointed out in v6, this condition will always be false, I suggest to drop 4/6 patch and this hunk here so it won't to confuse readers with assumption that condition might succeed. > + error_setg(errp, "%s device must be added as a child device before " > + "realize", TYPE_FW_CFG); > + return; > + } > > fw_cfg_add_bytes(s, FW_CFG_SIGNATURE, (char *)"QEMU", 4); > fw_cfg_add_bytes(s, FW_CFG_UUID, &qemu_uuid, 16); > @@ -965,7 +965,9 @@ FWCfgState *fw_cfg_init_io_dma(uint32_t iobase, uint32_t dma_iobase, > qdev_prop_set_bit(dev, "dma_enabled", false); > } > > - fw_cfg_init1(dev); > + object_property_add_child(OBJECT(qdev_get_machine()), TYPE_FW_CFG, > + OBJECT(dev), NULL); > + qdev_init_nofail(dev); > > sbd = SYS_BUS_DEVICE(dev); > ios = FW_CFG_IO(dev); > @@ -1003,7 +1005,9 @@ FWCfgState *fw_cfg_init_mem_wide(hwaddr ctl_addr, > qdev_prop_set_bit(dev, "dma_enabled", false); > } > > - fw_cfg_init1(dev); > + object_property_add_child(OBJECT(qdev_get_machine()), TYPE_FW_CFG, > + OBJECT(dev), NULL); > + qdev_init_nofail(dev); > > sbd = SYS_BUS_DEVICE(dev); > sysbus_mmio_map(sbd, 0, ctl_addr); > @@ -1033,6 +1037,7 @@ FWCfgState *fw_cfg_find(void) > return FW_CFG(object_resolve_path_type("", TYPE_FW_CFG, NULL)); > } > > + > static void fw_cfg_class_init(ObjectClass *klass, void *data) > { > DeviceClass *dc = DEVICE_CLASS(klass); > @@ -1104,6 +1109,12 @@ static void fw_cfg_io_realize(DeviceState *dev, Error **errp) > &fw_cfg_dma_mem_ops, FW_CFG(s), "fwcfg.dma", > sizeof(dma_addr_t)); > } > + > + fw_cfg_common_realize(dev, &local_err); > + if (local_err) { > + error_propagate(errp, local_err); > + return; > + } > } > > static void fw_cfg_io_class_init(ObjectClass *klass, void *data) > @@ -1170,6 +1181,12 @@ static void fw_cfg_mem_realize(DeviceState *dev, Error **errp) > sizeof(dma_addr_t)); > sysbus_init_mmio(sbd, &FW_CFG(s)->dma_iomem); > } > + > + fw_cfg_common_realize(dev, &local_err); > + if (local_err) { > + error_propagate(errp, local_err); > + return; > + } > } > > static void fw_cfg_mem_class_init(ObjectClass *klass, void *data) patch looks good, with above comments fixed: Reviewed-by: Igor Mammedov