Clean up use of qdev_init()
cover Clean up use of qdev_init()
qdev_init() can fail since commit 81a322d4. Most callers don't bother to check. This is less serious than it sounds, because they typically work with devices for which qdev_init() never fails. It's still unclean.
The last patch makes the compiler warn when the value of qdev_init() isn't used. If that warning triggers after merging this series, more unchecked calls have crept in since the base of this series (commit 2637c754). Feel free to ask me for a respin then.
Markus Armbruster (6):
Move function definitions out of xilinx.h
Make qdev_init() destroy the device on failure
Check return value of qdev_init()
New qdev_init_nofail()
Make isa_create() terminate program on failure
Warn if value of qdev_init() isn't checked
hw/apb_pci.c | 2 +- hw/arm_sysctl.c | 2 +- hw/armv7m.c | 6 ++-- hw/axis_dev88.c | 2 +- hw/escc.c | 4 +- hw/esp.c | 2 +- hw/etraxfs.c | 2 +- hw/fdc.c | 4 +- hw/grackle_pci.c | 2 +- hw/i2c.c | 2 +- hw/integratorcp.c | 2 +- hw/isa-bus.c | 11 +++------ hw/m48t59.c | 2 +- hw/mc146818rtc.c | 2 +- hw/mips_malta.c | 2 +- hw/musicpal.c | 4 +- hw/ne2000-isa.c | 2 +- hw/pc.c | 2 +- hw/pci-hotplug.c | 4 +- hw/pci.c | 9 ++++--- hw/piix_pci.c | 2 +- hw/qdev.c | 26 +++++++++++++++++++--- hw/qdev.h | 3 +- hw/scsi-bus.c | 4 ++- hw/smc91c111.c | 2 +- hw/ssi.c | 2 +- hw/stellaris.c | 2 +- hw/sun4m.c | 28 ++++++++++++------------ hw/sun4u.c | 4 +- hw/syborg.c | 4 +- hw/sysbus.c | 2 +- hw/unin_pci.c | 2 +- hw/usb-bus.c | 2 +- hw/usb-msd.c | 3 +- hw/vga-pci.c | 2 +- hw/xilinx.h | 58 ++++++++------------------------------------------ hw/xilinx_ethlite.c | 19 ++++++++++++++++ hw/xilinx_intc.c | 14 ++++++++++++ hw/xilinx_timer.c | 15 +++++++++++++ usb-linux.c | 4 +- 40 files changed, 148 insertions(+), 118 deletions(-)
PATCH 1/6 Move function definitions out of xilinx.h
xilinx.h defines a couple of static inline functions for creating devices. While that's a fair technique for hot functions, device initialization is about as cold as it gets. Define them in the device source files instead, and keep only declarations in the header.
--- hw/xilinx.h | 58 ++++++++------------------------------------------ hw/xilinx_ethlite.c | 19 ++++++++++++++++ hw/xilinx_intc.c | 14 ++++++++++++ hw/xilinx_timer.c | 15 +++++++++++++ 4 files changed, 58 insertions(+), 48 deletions(-)
hw/xilinx.h +10 -48
@@ -1,50 +1,12 @@ +#ifndef XILINX_H +#define XILINX_H -/* OPB Interrupt Controller. */ qemu_irq *microblaze_pic_init_cpu (CPUState *env); - -static inline DeviceState * -xilinx_intc_create (target_phys_addr_t base,qemu_irq irq,int kind_of_intr) -{ -DeviceState *dev; - - dev =qdev_create (NULL ,"xilinx,intc" ); -qdev_prop_set_uint32 (dev,"kind-of-intr" , kind_of_intr); -qdev_init (dev); -sysbus_mmio_map (sysbus_from_qdev (dev),0 , base); -sysbus_connect_irq (sysbus_from_qdev (dev),0 , irq); -return dev; -} - -/* OPB Timer/Counter. */ -static inline DeviceState * -xilinx_timer_create (target_phys_addr_t base,qemu_irq irq,int nr,int freq) -{ -DeviceState *dev; - - dev =qdev_create (NULL ,"xilinx,timer" ); -qdev_prop_set_uint32 (dev,"nr-timers" , nr); -qdev_prop_set_uint32 (dev,"frequency" , freq); -qdev_init (dev); -sysbus_mmio_map (sysbus_from_qdev (dev),0 , base); -sysbus_connect_irq (sysbus_from_qdev (dev),0 , irq); -return dev; -} - -/* XPS Ethernet Lite MAC. */ -static inline DeviceState * -xilinx_ethlite_create (NICInfo *nd,target_phys_addr_t base,qemu_irq irq, -int txpingpong,int rxpingpong) -{ -DeviceState *dev; - -qemu_check_nic_model (nd,"xilinx-ethlite" ); - - dev =qdev_create (NULL ,"xilinx,ethlite" ); - dev->nd = nd; -qdev_prop_set_uint32 (dev,"txpingpong" , txpingpong); -qdev_prop_set_uint32 (dev,"rxpingpong" , rxpingpong); -qdev_init (dev); -sysbus_mmio_map (sysbus_from_qdev (dev),0 , base); -sysbus_connect_irq (sysbus_from_qdev (dev),0 , irq); -return dev; -} +DeviceState *xilinx_intc_create (target_phys_addr_t base,qemu_irq irq, +int kind_of_intr); +DeviceState *xilinx_timer_create (target_phys_addr_t base,qemu_irq irq, +int nr,int freq); +DeviceState *xilinx_ethlite_create (NICInfo *nd,target_phys_addr_t base, +qemu_irq irq,int txpingpong,int rxpingpong); + +#endif
hw/xilinx_ethlite.c +19 -0
@@ -25,6 +25,7 @@#include "sysbus.h" #include "hw.h" #include "net.h" +#include "xilinx.h" #define D (x)#define R_TX_BUF0 0
@@ -201,6 +202,24 @@ static ssize_t eth_rx(VLANClientState *vc, const uint8_t *buf, size_t size)return size; } +DeviceState * +xilinx_ethlite_create (NICInfo *nd,target_phys_addr_t base,qemu_irq irq, +int txpingpong,int rxpingpong) +{ +DeviceState *dev; + +qemu_check_nic_model (nd,"xilinx-ethlite" ); + + dev =qdev_create (NULL ,"xilinx,ethlite" ); + dev->nd = nd; +qdev_prop_set_uint32 (dev,"txpingpong" , txpingpong); +qdev_prop_set_uint32 (dev,"rxpingpong" , rxpingpong); +qdev_init (dev); +sysbus_mmio_map (sysbus_from_qdev (dev),0 , base); +sysbus_connect_irq (sysbus_from_qdev (dev),0 , irq); +return dev; +} +static void eth_cleanup (VLANClientState *vc) {struct xlx_ethlite *s = vc->opaque;
hw/xilinx_intc.c +14 -0
@@ -24,6 +24,7 @@#include "sysbus.h" #include "hw.h" +#include "xilinx.h" #define D (x)
@@ -145,6 +146,19 @@ static void irq_handler(void *opaque, int irq, int level)update_irq (p); } +DeviceState * +xilinx_intc_create (target_phys_addr_t base,qemu_irq irq,int kind_of_intr) +{ +DeviceState *dev; + + dev =qdev_create (NULL ,"xilinx,intc" ); +qdev_prop_set_uint32 (dev,"kind-of-intr" , kind_of_intr); +qdev_init (dev); +sysbus_mmio_map (sysbus_from_qdev (dev),0 , base); +sysbus_connect_irq (sysbus_from_qdev (dev),0 , irq); +return dev; +} +static int xilinx_intc_init (SysBusDevice *dev) {struct xlx_pic *p =FROM_SYSBUS (typeof (*p), dev);
hw/xilinx_timer.c +15 -0
@@ -25,6 +25,7 @@#include "sysbus.h" #include "sysemu.h" #include "qemu-timer.h" +#include "xilinx.h" #define D (x)
@@ -189,6 +190,20 @@ static void timer_hit(void *opaque)timer_update_irq (t); } +DeviceState * +xilinx_timer_create (target_phys_addr_t base,qemu_irq irq,int nr,int freq) +{ +DeviceState *dev; + + dev =qdev_create (NULL ,"xilinx,timer" ); +qdev_prop_set_uint32 (dev,"nr-timers" , nr); +qdev_prop_set_uint32 (dev,"frequency" , freq); +qdev_init (dev); +sysbus_mmio_map (sysbus_from_qdev (dev),0 , base); +sysbus_connect_irq (sysbus_from_qdev (dev),0 , irq); +return dev; +} +static int xilinx_timer_init (SysBusDevice *dev) {struct timerblock *t =FROM_SYSBUS (typeof (*t), dev);
Signature · 1 line
1.6.2.5
Re: [Qemu-devel] [PATCH 1/6] Move function definitions out of xilinx.h
Quoted 4 lines · Markus Armbruster
xilinx.hdefines a couple of static inline functions for creating devices. While that's a fair technique for hot functions, device initialization is about as cold as it gets. Define them in the device source files instead, and keep only declarations in the header.
If I understood the qdev plan correctly, this is going to wrong direction. These functions should reside near the instantiation, not in the device code. The current approach looks OK if there are going to be more users of the devices.
Re: [Qemu-devel] [PATCH 1/6] Move function definitions out of xilinx.h
Quoted 10 lines · Blue Swirl
Quoted 4 lines · Markus Armbruster
xilinx.hdefines a couple of static inline functions for creating devices. While that's a fair technique for hot functions, device initialization is about as cold as it gets. Define them in the device source files instead, and keep only declarations in the header.If I understood the qdev plan correctly, this is going to wrong direction. These functions should reside near the instantiation, not in the device code. The current approach looks OK if there are going to be more users of the devices.
I agree that they shouldn't be in the device source.
The reason they ended up in a header and not with the petalogix board was that in my tree there are multiple boards using these functions to easy instantiate devices.
Cheers
Re: [Qemu-devel] [PATCH 1/6] Move function definitions out of xilinx.h
Quoted 11 lines · Edgar E. Iglesias
Quoted 10 lines · Blue Swirl
Quoted 4 lines · Markus Armbruster
xilinx.hdefines a couple of static inline functions for creating devices. While that's a fair technique for hot functions, device initialization is about as cold as it gets. Define them in the device source files instead, and keep only declarations in the header.If I understood the qdev plan correctly, this is going to wrong direction. These functions should reside near the instantiation, not in the device code. The current approach looks OK if there are going to be more users of the devices.
The functions should go away ;)
Some day the information carried by those code snippeds should come from a machine description file, then we'll don't need them any more.
Quoted 4 lines
I agree that they shouldn't be in the device source.
The reason they ended up in a header and not with the petalogix board was that in my tree there are multiple boards using these functions to easy instantiate devices.
They have to be somewhere. Having them in a header file is unclean. Having them in the board-specific code isn't practical when multiple boards share the code. I'd stick them to the device source code as well. Also note that this is common practice elsewhere in the tree.
cheers
Gerd
Re: [Qemu-devel] [PATCH 1/6] Move function definitions out of xilinx.h
Quoted 14 lines · Gerd Hoffmann
Quoted 8 lines · Edgar E. Iglesias
Quoted 7 lines · Blue Swirl
Quoted 4 lines · Markus Armbruster
xilinx.hdefines a couple of static inline functions for creating devices. While that's a fair technique for hot functions, device initialization is about as cold as it gets. Define them in the device source files instead, and keep only declarations in the header.Quoted 4 lines
Quoted 4 lines
If I understood the qdev plan correctly, this is going to wrong direction. These functions should reside near the instantiation, not in the device code. The current approach looks OK if there are going to be more users of the devices.
Quoted 4 lines
The functions should go away ;)
Some day the information carried by those code snippeds should come from a machine description file, then we'll don't need them any more.
I agree.
Quoted 10 lines
Quoted 5 lines
I agree that they shouldn't be in the device source.
The reason they ended up in a header and not with the petalogix board was that in my tree there are multiple boards using these functions to easy instantiate devices.They have to be somewhere. Having them in a header file is unclean. Having them in the board-specific code isn't practical when multiple boards share the code. I'd stick them to the device source code as well. Also note that this is common practice elsewhere in the tree.
I disagree.
But if ppl feel very strongly about this, I can remove them and deal with the code duplication in my tree. Afterall, it's unlikely that upsteam qemu gets more xilinx boards before some kind of device tree driven board support is there.
Cheers
Re: [Qemu-devel] [PATCH 1/6] Move function definitions out of xilinx.h
Quoted 33 lines · Edgar E. Iglesias
Quoted 14 lines · Gerd Hoffmann
Quoted 8 lines · Edgar E. Iglesias
Quoted 7 lines · Blue Swirl
Quoted 4 lines · Markus Armbruster
xilinx.hdefines a couple of static inline functions for creating devices. While that's a fair technique for hot functions, device initialization is about as cold as it gets. Define them in the device source files instead, and keep only declarations in the header.Quoted 4 lines
Quoted 4 lines
If I understood the qdev plan correctly, this is going to wrong direction. These functions should reside near the instantiation, not in the device code. The current approach looks OK if there are going to be more users of the devices.
Quoted 4 lines
The functions should go away ;)
Some day the information carried by those code snippeds should come from a machine description file, then we'll don't need them any more.
I agree.
Quoted 10 lines
Quoted 5 lines
I agree that they shouldn't be in the device source.
The reason they ended up in a header and not with the petalogix board was that in my tree there are multiple boards using these functions to easy instantiate devices.They have to be somewhere. Having them in a header file is unclean. Having them in the board-specific code isn't practical when multiple boards share the code. I'd stick them to the device source code as well. Also note that this is common practice elsewhere in the tree.
Quoted 6 lines
I disagree.
But if ppl feel very strongly about this, I can remove them and deal with the code duplication in my tree. Afterall, it's unlikely that upsteam qemu gets more xilinx boards before some kind of device tree driven board support is there.
I also disagree. Why would the functions in a header file be unclean? The common practice is wrong.
Re: [Qemu-devel] [PATCH 1/6] Move function definitions out of xilinx.h
Quoted 37 lines · Blue Swirl
Quoted 35 lines · Edgar E. Iglesias
Quoted 14 lines · Gerd Hoffmann
Quoted 8 lines · Edgar E. Iglesias
Quoted 7 lines · Blue Swirl
Quoted 4 lines · Markus Armbruster
xilinx.hdefines a couple of static inline functions for creating devices. While that's a fair technique for hot functions, device initialization is about as cold as it gets. Define them in the device source files instead, and keep only declarations in the header.Quoted 4 lines
Quoted 4 lines
If I understood the qdev plan correctly, this is going to wrong direction. These functions should reside near the instantiation, not in the device code. The current approach looks OK if there are going to be more users of the devices.
Quoted 4 lines
The functions should go away ;)
Some day the information carried by those code snippeds should come from a machine description file, then we'll don't need them any more.
I agree.
Quoted 10 lines
Quoted 5 lines
I agree that they shouldn't be in the device source.
The reason they ended up in a header and not with the petalogix board was that in my tree there are multiple boards using these functions to easy instantiate devices.They have to be somewhere. Having them in a header file is unclean. Having them in the board-specific code isn't practical when multiple boards share the code. I'd stick them to the device source code as well. Also note that this is common practice elsewhere in the tree.
I disagree.
Quoted 7 lines
Quoted 4 lines
But if ppl feel very strongly about this, I can remove them and deal with the code duplication in my tree. Afterall, it's unlikely that upsteam qemu gets more xilinx boards before some kind of device tree driven board support is there.
I also disagree. Why would the functions in a header file be unclean? The common practice is wrong.
The common practice is to put code in .c files, and stick to declarations needed by several .c files in .h files. It's okay to deviate from that common practice if there's a good reason --- we do that in places. However, in this case, I couldn't see a reason, so I moved the code from the header to the place where such code lives for pretty much every other qdevified device: the device code.
Doing it the same way everywhere means that maintainers have an easier time to predict where stuff is, how it's named, and so forth. Don't disappoint reasonable maintainer expectations without need.
I don't want to make a big deal out of this. If the code must remain in the header, expectations be damned, I'll respin the series and move on.
Re: [Qemu-devel] [PATCH 1/6] Move function definitions out of xilinx.h
Quoted 47 lines · Markus Armbruster
Quoted 37 lines · Blue Swirl
Quoted 35 lines · Edgar E. Iglesias
Quoted 14 lines · Gerd Hoffmann
Quoted 8 lines · Edgar E. Iglesias
Quoted 7 lines · Blue Swirl
Quoted 4 lines · Markus Armbruster
xilinx.hdefines a couple of static inline functions for creating devices. While that's a fair technique for hot functions, device initialization is about as cold as it gets. Define them in the device source files instead, and keep only declarations in the header.Quoted 4 lines
Quoted 4 lines
If I understood the qdev plan correctly, this is going to wrong direction. These functions should reside near the instantiation, not in the device code. The current approach looks OK if there are going to be more users of the devices.
Quoted 4 lines
The functions should go away ;)
Some day the information carried by those code snippeds should come from a machine description file, then we'll don't need them any more.
I agree.
Quoted 10 lines
Quoted 5 lines
I agree that they shouldn't be in the device source.
The reason they ended up in a header and not with the petalogix board was that in my tree there are multiple boards using these functions to easy instantiate devices.They have to be somewhere. Having them in a header file is unclean. Having them in the board-specific code isn't practical when multiple boards share the code. I'd stick them to the device source code as well. Also note that this is common practice elsewhere in the tree.
I disagree.
Quoted 7 lines
Quoted 4 lines
But if ppl feel very strongly about this, I can remove them and deal with the code duplication in my tree. Afterall, it's unlikely that upsteam qemu gets more xilinx boards before some kind of device tree driven board support is there.
I also disagree. Why would the functions in a header file be unclean? The common practice is wrong.
Quoted 6 lines
The common practice is to put code in .c files, and stick to declarations needed by several .c files in .h files. It's okay to deviate from that common practice if there's a good reason --- we do that in places. However, in this case, I couldn't see a reason, so I moved the code from the header to the place where such code lives for pretty much every other qdevified device: the device code.
Understandable, since we don't see the source code of the other xilinx boards.
Quoted 3 lines
Doing it the same way everywhere means that maintainers have an easier time to predict where stuff is, how it's named, and so forth. Don't disappoint reasonable maintainer expectations without need.
It's still wrong, however disappointing that may be. But with enough cleanups and guidance, the expectations will be directed to right path.
Quoted 2 lines
I don't want to make a big deal out of this. If the code must remain in the header, expectations be damned, I'll respin the series and move on.
Pushing the device instantiation code higher makes the device code a bit simpler. Also the internal state can be kept private, provided that the interface is designed properly. For example, slavio_timer.c does not export internal state and all functions are static. The former init function is now in sun4m.c.
Actually, while I'm quite happy with the device level design (and I think other targets should also be worked to this direction, not the opposite), the next level design is not so clear since we don't have the machine config or something.
PATCH 2/6 Make qdev_init() destroy the device on failure
In reply to Markus ArmbrusterBefore, every caller had to do this. Only two actually did.
--- hw/isa-bus.c | 1 - hw/qdev.c | 11 +++++++---- 2 files changed, 7 insertions(+), 5 deletions(-)
hw/isa-bus.c +0 -1
@@ -127,7 +127,6 @@ ISADevice *isa_create_simple(const char *name) dev =isa_create (name);if (qdev_init (&dev->qdev) !=0 ) { -qdev_free (&dev->qdev);return NULL ; }return dev;
hw/qdev.c +7 -4
@@ -202,8 +202,7 @@ DeviceState *qdev_device_add(QemuOpts *opts)qdev_free (qdev);return NULL ; } -if (qdev_init (qdev) !=0 ) { -qdev_free (qdev); +if (qdev_init (qdev) <0 ) {return NULL ; }return qdev;
@@ -211,14 +210,18 @@ DeviceState *qdev_device_add(QemuOpts *opts)/* Initialize a device. Device properties should be set before calling this function. IRQs and MMIO regions should be connected/mapped after -calling this function. */ +calling this function. +On failure, destroy the device and return negative value. +Return 0 on success. */ int qdev_init (DeviceState *dev) {int rc; rc = dev->info->init (dev, dev->info); -if (rc <0 ) +if (rc <0 ) { +qdev_free (dev);return rc; + }if (dev->info->reset)qemu_register_reset (dev->info->reset, dev);if (dev->info->vmsd)
Signature · 1 line
1.6.2.5
PATCH 3/6 Check return value of qdev_init()
In reply to Markus ArmbrusterBut do so only where it may actually fail. Leave the rest for the next commit.
--- hw/pci-hotplug.c | 4 ++-- hw/pci.c | 4 +++- hw/scsi-bus.c | 4 +++- hw/usb-msd.c | 3 ++- 4 files changed, 10 insertions(+), 5 deletions(-)
hw/pci-hotplug.c +2 -2
@@ -149,8 +149,8 @@ static PCIDevice *qemu_pci_hot_add_storage(Monitor *mon,default : dev =NULL ; } -if (dev) -qdev_init (&dev->qdev); +if (!dev ||qdev_init (&dev->qdev) <0 ) +return NULL ;return dev; }
hw/pci.c +3 -1
@@ -807,6 +807,7 @@ static const char * const pci_nic_names[] = { };/* Initialize a PCI NIC. */ +/* FIXME callers should check for failure, but don't */ PCIDevice *pci_nic_init (NICInfo *nd,const char *default_model,const char *default_devaddr) {
@@ -824,7 +825,8 @@ PCIDevice *pci_nic_init(NICInfo *nd, const char *default_model,if (nd->id) dev->id =qemu_strdup (nd->id); dev->nd = nd; -qdev_init (dev); +if (qdev_init (dev) <0 ) +return NULL ; nd->private = dev;return pci_dev; }
hw/scsi-bus.c +3 -1
@@ -65,6 +65,7 @@ void scsi_qdev_register(SCSIDeviceInfo *info) }/* handle legacy '-drive if=scsi,...' cmd line args */ +/* FIXME callers should check for failure, but don't */ SCSIDevice *scsi_bus_legacy_add_drive (SCSIBus *bus,DriveInfo *dinfo,int unit) {const char *driver;
@@ -74,7 +75,8 @@ SCSIDevice *scsi_bus_legacy_add_drive(SCSIBus *bus, DriveInfo *dinfo, int unit) dev =qdev_create (&bus->qbus, driver);qdev_prop_set_uint32 (dev,"scsi-id" , unit);qdev_prop_set_drive (dev,"drive" , dinfo); -qdev_init (dev); +if (qdev_init (dev) <0 ) +return NULL ;return DO_UPCAST (SCSIDevice, qdev, dev); }
hw/usb-msd.c +2 -1
@@ -579,7 +579,8 @@ USBDevice *usb_msd_init(const char *filename)/* create guest device */ dev =usb_create (NULL /* FIXME */ ,"QEMU USB MSD" );qdev_prop_set_drive (&dev->qdev,"drive" , dinfo); -qdev_init (&dev->qdev); +if (qdev_init (&dev->qdev) <0 ) +return NULL ;return dev; }
Signature · 1 line
1.6.2.5
PATCH 4/6 New qdev_init_nofail()
In reply to Markus ArmbrusterLike qdev_init(), but terminate program via hw_error() instead of returning an error value.
Use it instead of qdev_init() where terminating the program on failure is okay, either because it's during machine construction, or because we know that failure can't happen.
Because relying in the latter is somewhat unclean, and the former is not always obvious, it would be nice to go back to qdev_init() in the not-so-obvious cases, only with proper error handling. I'm leaving that for another day, because it involves making sure that error values are properly checked by all callers.
--- hw/apb_pci.c | 2 +- hw/arm_sysctl.c | 2 +- hw/armv7m.c | 6 +++--- hw/axis_dev88.c | 2 +- hw/escc.c | 4 ++-- hw/esp.c | 2 +- hw/etraxfs.c | 2 +- hw/fdc.c | 4 ++-- hw/grackle_pci.c | 2 +- hw/i2c.c | 2 +- hw/integratorcp.c | 2 +- hw/isa-bus.c | 2 +- hw/m48t59.c | 2 +- hw/mc146818rtc.c | 2 +- hw/mips_malta.c | 2 +- hw/musicpal.c | 4 ++-- hw/ne2000-isa.c | 2 +- hw/pc.c | 2 +- hw/pci.c | 5 ++--- hw/piix_pci.c | 2 +- hw/qdev.c | 15 +++++++++++++++ hw/qdev.h | 1 + hw/smc91c111.c | 2 +- hw/ssi.c | 2 +- hw/stellaris.c | 2 +- hw/sun4m.c | 28 ++++++++++++++-------------- hw/sun4u.c | 4 ++-- hw/syborg.c | 4 ++-- hw/sysbus.c | 2 +- hw/unin_pci.c | 2 +- hw/usb-bus.c | 2 +- hw/vga-pci.c | 2 +- hw/xilinx_ethlite.c | 2 +- hw/xilinx_intc.c | 2 +- hw/xilinx_timer.c | 2 +- usb-linux.c | 4 ++-- 36 files changed, 72 insertions(+), 57 deletions(-)
hw/apb_pci.c +1 -1
@@ -236,7 +236,7 @@ PCIBus *pci_apb_init(target_phys_addr_t special_base,/* Ultrasparc PBM main bus */ dev =qdev_create (NULL ,"pbm" ); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);/* apb_config */ sysbus_mmio_map (s,0 , special_base +0x2000ULL );
hw/arm_sysctl.c +1 -1
@@ -211,7 +211,7 @@ void arm_sysctl_init(uint32_t base, uint32_t sys_id) dev =qdev_create (NULL ,"realview_sysctl" );qdev_prop_set_uint32 (dev,"sys_id" , sys_id); -qdev_init (dev); +qdev_init_nofail (dev);sysbus_mmio_map (sysbus_from_qdev (dev),0 , base); }
hw/armv7m.c +3 -3
@@ -139,12 +139,12 @@ static void armv7m_bitband_init(void) dev =qdev_create (NULL ,"ARM,bitband-memory" );qdev_prop_set_uint32 (dev,"base" ,0x20000000 ); -qdev_init (dev); +qdev_init_nofail (dev);sysbus_mmio_map (sysbus_from_qdev (dev),0 ,0x22000000 ); dev =qdev_create (NULL ,"ARM,bitband-memory" );qdev_prop_set_uint32 (dev,"base" ,0x40000000 ); -qdev_init (dev); +qdev_init_nofail (dev);sysbus_mmio_map (sysbus_from_qdev (dev),0 ,0x42000000 ); }
@@ -199,7 +199,7 @@ qemu_irq *armv7m_init(int flash_size, int sram_size, nvic =qdev_create (NULL ,"armv7m_nvic" ); env->v7m.nvic = nvic; -qdev_init (nvic); +qdev_init_nofail (nvic); cpu_pic =arm_pic_init_cpu (env);sysbus_connect_irq (sysbus_from_qdev (nvic),0 , cpu_pic[ARM_PIC_CPU_IRQ ]);for (i =0 ; i <64 ; i++) {
hw/axis_dev88.c +1 -1
@@ -298,7 +298,7 @@ void axisdev88_init (ram_addr_t ram_size, dev =qdev_create (NULL ,"etraxfs,pic" );/* FIXME: Is there a proper way to signal vectors to the CPU core? */ qdev_prop_set_ptr (dev,"interrupt_vector" , &env->interrupt_vector); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);sysbus_mmio_map (s,0 ,0x3001c000 );sysbus_connect_irq (s,0 , cpu_irq[0 ]);
hw/escc.c +2 -2
@@ -741,7 +741,7 @@ int escc_init(target_phys_addr_t base, qemu_irq irqA, qemu_irq irqB,qdev_prop_set_chr (dev,"chrA" , chrA);qdev_prop_set_uint32 (dev,"chnBtype" , ser);qdev_prop_set_uint32 (dev,"chnAtype" , ser); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);sysbus_connect_irq (s,0 , irqA);sysbus_connect_irq (s,1 , irqB);
@@ -904,7 +904,7 @@ void slavio_serial_ms_kbd_init(target_phys_addr_t base, qemu_irq irq,qdev_prop_set_chr (dev,"chrA" ,NULL );qdev_prop_set_uint32 (dev,"chnBtype" , mouse);qdev_prop_set_uint32 (dev,"chnAtype" , kbd); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);sysbus_connect_irq (s,0 , irq);sysbus_connect_irq (s,1 , irq);
hw/esp.c +1 -1
@@ -667,7 +667,7 @@ void esp_init(target_phys_addr_t espaddr, int it_shift, esp->dma_memory_write = dma_memory_write; esp->dma_opaque = dma_opaque; esp->it_shift = it_shift; -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);sysbus_connect_irq (s,0 , irq);sysbus_mmio_map (s,0 , espaddr);
hw/etraxfs.c +1 -1
@@ -90,7 +90,7 @@ void bareetraxfs_init (ram_addr_t ram_size, dev =qdev_create (NULL ,"etraxfs,pic" );/* FIXME: Is there a proper way to signal vectors to the CPU core? */ qdev_prop_set_ptr (dev,"interrupt_vector" , &env->interrupt_vector); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);sysbus_mmio_map (s,0 ,0x3001c000 );sysbus_connect_irq (s,0 , cpu_irq[0 ]);
hw/fdc.c +2 -2
@@ -1865,7 +1865,7 @@ fdctrl_t *fdctrl_init_sysbus(qemu_irq irq, int dma_chann,fdctrl_sysbus_t *sys; dev =qdev_create (NULL ,"sysbus-fdc" ); -qdev_init (dev); +qdev_init_nofail (dev); sys =DO_UPCAST (fdctrl_sysbus_t, busdev.qdev, dev); fdctrl = &sys->state;sysbus_connect_irq (&sys->busdev,0 , irq);
@@ -1886,7 +1886,7 @@ fdctrl_t *sun4m_fdctrl_init (qemu_irq irq, target_phys_addr_t io_base,fdctrl_t *fdctrl; dev =qdev_create (NULL ,"SUNW,fdtwo" ); -qdev_init (dev); +qdev_init_nofail (dev); sys =DO_UPCAST (fdctrl_sysbus_t, busdev.qdev, dev); fdctrl = &sys->state;sysbus_connect_irq (&sys->busdev,0 , irq);
hw/grackle_pci.c +1 -1
@@ -138,7 +138,7 @@ PCIBus *pci_grackle_init(uint32_t base, qemu_irq *pic)GrackleState *d; dev =qdev_create (NULL ,"grackle" ); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev); d =FROM_SYSBUS (GrackleState, s); d->host_state.bus = pci_register_bus(NULL, "pci",
hw/i2c.c +1 -1
@@ -173,6 +173,6 @@ DeviceState *i2c_create_slave(i2c_bus *bus, const char *name, int addr) dev =qdev_create (&bus->qbus, name);qdev_prop_set_uint32 (dev,"address" , addr); -qdev_init (dev); +qdev_init_nofail (dev);return dev; }
hw/integratorcp.c +1 -1
@@ -477,7 +477,7 @@ static void integratorcp_init(ram_addr_t ram_size, dev =qdev_create (NULL ,"integrator_core" );qdev_prop_set_uint32 (dev,"memsz" , ram_size >>20 ); -qdev_init (dev); +qdev_init_nofail (dev);sysbus_mmio_map ((SysBusDevice *)dev,0 ,0x10000000 ); cpu_pic =arm_pic_init_cpu (env);
hw/isa-bus.c +1 -1
@@ -45,7 +45,7 @@ ISABus *isa_bus_new(DeviceState *dev) }if (NULL == dev) { dev =qdev_create (NULL ,"isabus-bridge" ); -qdev_init (dev); +qdev_init_nofail (dev); } isabus =FROM_QBUS (ISABus,qbus_create (&isa_bus_info, dev,NULL ));
hw/m48t59.c +1 -1
@@ -626,7 +626,7 @@ m48t59_t *m48t59_init (qemu_irq IRQ, target_phys_addr_t mem_base,qdev_prop_set_uint32 (dev,"type" , type);qdev_prop_set_uint32 (dev,"size" , size);qdev_prop_set_uint32 (dev,"io_base" , io_base); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);sysbus_connect_irq (s,0 ,IRQ );if (io_base !=0 ) {
hw/mc146818rtc.c +1 -1
@@ -637,7 +637,7 @@ RTCState *rtc_init(int base_year) dev =isa_create ("mc146818rtc" );qdev_prop_set_int32 (&dev->qdev,"base_year" , base_year); -qdev_init (&dev->qdev); +qdev_init_nofail (&dev->qdev);return DO_UPCAST (RTCState, dev, dev); }
hw/mips_malta.c +1 -1
@@ -915,7 +915,7 @@ void mips_malta_init (ram_addr_t ram_size, eeprom =qdev_create ((BusState *)smbus,"smbus-eeprom" );qdev_prop_set_uint32 (eeprom,"address" ,0x50 + i);qdev_prop_set_ptr (eeprom,"data" , eeprom_buf + (i *256 )); -qdev_init (eeprom); +qdev_init_nofail (eeprom); } pit =pit_init (0x40 ,isa_reserve_irq (0 ));DMA_init (0 );
hw/musicpal.c +2 -2
@@ -1357,7 +1357,7 @@ static void musicpal_init(ram_addr_t ram_size,qemu_check_nic_model (&nd_table[0 ],"mv88w8618" ); dev =qdev_create (NULL ,"mv88w8618_eth" ); dev->nd = &nd_table[0 ]; -qdev_init (dev); +qdev_init_nofail (dev);sysbus_mmio_map (sysbus_from_qdev (dev),0 ,MP_ETH_BASE );sysbus_connect_irq (sysbus_from_qdev (dev),0 , pic[MP_ETH_IRQ ]);
@@ -1390,7 +1390,7 @@ static void musicpal_init(ram_addr_t ram_size, dev =qdev_create (NULL ,"mv88w8618_audio" ); s =sysbus_from_qdev (dev);qdev_prop_set_ptr (dev,"wm8750" , wm8750_dev); -qdev_init (dev); +qdev_init_nofail (dev);sysbus_mmio_map (s,0 ,MP_AUDIO_BASE );sysbus_connect_irq (s,0 , pic[MP_AUDIO_IRQ ]);#endif
hw/ne2000-isa.c +1 -1
@@ -89,7 +89,7 @@ void isa_ne2000_init(int base, int irq, NICInfo *nd) dev->qdev.nd = nd;/* hack alert */ qdev_prop_set_uint32 (&dev->qdev,"iobase" , base);qdev_prop_set_uint32 (&dev->qdev,"irq" , irq); -qdev_init (&dev->qdev); +qdev_init_nofail (&dev->qdev); }static ISADeviceInfo ne2000_isa_info = {
hw/pc.c +1 -1
@@ -1403,7 +1403,7 @@ static void pc_init1(ram_addr_t ram_size, eeprom =qdev_create ((BusState *)smbus,"smbus-eeprom" );qdev_prop_set_uint32 (eeprom,"address" ,0x50 + i);qdev_prop_set_ptr (eeprom,"data" , eeprom_buf + (i *256 )); -qdev_init (eeprom); +qdev_init_nofail (eeprom); } }
hw/pci.c +2 -3
@@ -931,9 +931,8 @@ PCIDevice *pci_create_simple(PCIBus *bus, int devfn, const char *name) dev =qdev_create (&bus->qbus, name);qdev_prop_set_uint32 (dev,"addr" , devfn); -qdev_init (dev); - -return (PCIDevice *)dev; +qdev_init_nofail (dev); +return DO_UPCAST (PCIDevice, qdev, dev); }static int pci_find_space (PCIDevice *pdev,uint8_t size)
hw/piix_pci.c +1 -1
@@ -240,7 +240,7 @@ PCIBus *i440fx_init(PCII440FXState **pi440fx_state, int *piix3_devfn, qemu_irq * b =pci_register_bus (&s->busdev.qdev,"pci.0" , piix3_set_irq, pci_slot_get_pirq, irq_state,0 ,4 ); s->bus = b; -qdev_init (dev); +qdev_init_nofail (dev); d =pci_create_simple (b,0 ,"i440FX" ); *pi440fx_state =DO_UPCAST (PCII440FXState, dev, d);
hw/qdev.c +15 -0
@@ -229,6 +229,21 @@ int qdev_init(DeviceState *dev)return 0 ; } +/* Like qdev_init(), but terminate program via hw_error() instead of +returning an error value. This is okay during machine creation. +Don't use for hotplug, because there callers need to recover from +failure. Exception: if you know the device's init() callback can't +fail, then qdev_init_nofail() can't fail either, and is therefore +usable even then. But relying on the device implementation that +way is somewhat unclean, and best avoided. */ +void qdev_init_nofail (DeviceState *dev) +{ +DeviceInfo *info = dev->info; + +if (qdev_init (dev) <0 ) +hw_error ("Initialization of device %s failed\n" , info->name); +} +/* Unlink device from bus and free the structure. */ void qdev_free (DeviceState *dev) {
hw/qdev.h +1 -0
@@ -90,6 +90,7 @@ struct CompatProperty {DeviceState *qdev_create (BusState *bus,const char *name);DeviceState *qdev_device_add (QemuOpts *opts);int qdev_init (DeviceState *dev); +void qdev_init_nofail (DeviceState *dev);void qdev_free (DeviceState *dev);qemu_irq qdev_get_gpio_in (DeviceState *dev,int n);
hw/smc91c111.c +1 -1
@@ -735,7 +735,7 @@ void smc91c111_init(NICInfo *nd, uint32_t base, qemu_irq irq)qemu_check_nic_model (nd,"smc91c111" ); dev =qdev_create (NULL ,"smc91c111" ); dev->nd = nd; -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);sysbus_mmio_map (s,0 , base);sysbus_connect_irq (s,0 , irq);
hw/ssi.c +1 -1
@@ -46,7 +46,7 @@ DeviceState *ssi_create_slave(SSIBus *bus, const char *name) {DeviceState *dev; dev =qdev_create (&bus->qbus, name); -qdev_init (dev); +qdev_init_nofail (dev);return dev; }
hw/stellaris.c +1 -1
@@ -1384,7 +1384,7 @@ static void stellaris_init(const char *kernel_filename, const char *cpu_model, enet =qdev_create (NULL ,"stellaris_enet" ); enet->nd = &nd_table[0 ]; -qdev_init (enet); +qdev_init_nofail (enet);sysbus_mmio_map (sysbus_from_qdev (enet),0 ,0x40048000 );sysbus_connect_irq (sysbus_from_qdev (enet),0 , pic[42 ]); }
hw/sun4m.c +14 -14
@@ -349,7 +349,7 @@ static void *iommu_init(target_phys_addr_t addr, uint32_t version, qemu_irq irq) dev =qdev_create (NULL ,"iommu" );qdev_prop_set_uint32 (dev,"version" , version); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);sysbus_connect_irq (s,0 , irq);sysbus_mmio_map (s,0 , addr);
@@ -365,7 +365,7 @@ static void *sparc32_dma_init(target_phys_addr_t daddr, qemu_irq parent_irq, dev =qdev_create (NULL ,"sparc32_dma" );qdev_prop_set_ptr (dev,"iommu_opaque" , iommu); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);sysbus_connect_irq (s,0 , parent_irq); *dev_irq =qdev_get_gpio_in (dev,0 );
@@ -386,7 +386,7 @@ static void lance_init(NICInfo *nd, target_phys_addr_t leaddr, dev =qdev_create (NULL ,"lance" ); dev->nd = nd;qdev_prop_set_ptr (dev,"dma" , dma_opaque); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);sysbus_mmio_map (s,0 , leaddr);sysbus_connect_irq (s,0 , irq);
@@ -403,7 +403,7 @@ static DeviceState *slavio_intctl_init(target_phys_addr_t addr,unsigned int i, j; dev =qdev_create (NULL ,"slavio_intctl" ); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);
@@ -432,7 +432,7 @@ static void slavio_timer_init_all(target_phys_addr_t addr, qemu_irq master_irq, dev =qdev_create (NULL ,"slavio_timer" );qdev_prop_set_uint32 (dev,"num_cpus" , num_cpus); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);sysbus_connect_irq (s,0 , master_irq);sysbus_mmio_map (s,0 , addr +SYS_TIMER_OFFSET );
@@ -458,7 +458,7 @@ static void slavio_misc_init(target_phys_addr_t base,SysBusDevice *s; dev =qdev_create (NULL ,"slavio_misc" ); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);if (base) {/* 8 bit registers */
@@ -495,7 +495,7 @@ static void ecc_init(target_phys_addr_t base, qemu_irq irq, uint32_t version) dev =qdev_create (NULL ,"eccmemctl" );qdev_prop_set_uint32 (dev,"version" , version); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);sysbus_connect_irq (s,0 , irq);sysbus_mmio_map (s,0 , base);
@@ -510,7 +510,7 @@ static void apc_init(target_phys_addr_t power_base, qemu_irq cpu_halt)SysBusDevice *s; dev =qdev_create (NULL ,"apc" ); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);/* Power management (APC) XXX: not a Slavio device */ sysbus_mmio_map (s,0 , power_base);
@@ -529,7 +529,7 @@ static void tcx_init(target_phys_addr_t addr, int vram_size, int width,qdev_prop_set_uint16 (dev,"width" , width);qdev_prop_set_uint16 (dev,"height" , height);qdev_prop_set_uint16 (dev,"depth" , depth); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);/* 8-bit plane */ sysbus_mmio_map (s,0 , addr +0x00800000ULL );
@@ -559,7 +559,7 @@ static void idreg_init(target_phys_addr_t addr)SysBusDevice *s; dev =qdev_create (NULL ,"macio_idreg" ); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);sysbus_mmio_map (s,0 , addr);
@@ -597,7 +597,7 @@ static void prom_init(target_phys_addr_t addr, const char *bios_name)int ret; dev =qdev_create (NULL ,"openprom" ); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);sysbus_mmio_map (s,0 , addr);
@@ -686,7 +686,7 @@ static void ram_init(target_phys_addr_t addr, ram_addr_t RAM_size, d =FROM_SYSBUS (RamDevice, s); d->size = RAM_size; -qdev_init (dev); +qdev_init_nofail (dev);sysbus_mmio_map (s,0 , addr); }
@@ -1337,7 +1337,7 @@ static DeviceState *sbi_init(target_phys_addr_t addr, qemu_irq **parent_irq)unsigned int i; dev =qdev_create (NULL ,"sbi" ); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);
@@ -1527,7 +1527,7 @@ static DeviceState *sun4c_intctl_init(target_phys_addr_t addr,unsigned int i; dev =qdev_create (NULL ,"sun4c_intctl" ); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);
hw/sun4u.c +2 -2
@@ -407,7 +407,7 @@ static void prom_init(target_phys_addr_t addr, const char *bios_name)int ret; dev =qdev_create (NULL ,"openprom" ); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);sysbus_mmio_map (s,0 , addr);
@@ -489,7 +489,7 @@ static void ram_init(target_phys_addr_t addr, ram_addr_t RAM_size) d =FROM_SYSBUS (RamDevice, s); d->size = RAM_size; -qdev_init (dev); +qdev_init_nofail (dev);sysbus_mmio_map (s,0 , addr); }
hw/syborg.c +2 -2
@@ -65,7 +65,7 @@ static void syborg_init(ram_addr_t ram_size, dev =qdev_create (NULL ,"syborg,timer" );qdev_prop_set_uint32 (dev,"frequency" ,1000000 ); -qdev_init (dev); +qdev_init_nofail (dev);sysbus_mmio_map (sysbus_from_qdev (dev),0 ,0xC0002000 );sysbus_connect_irq (sysbus_from_qdev (dev),0 , pic[1 ]);
@@ -84,7 +84,7 @@ static void syborg_init(ram_addr_t ram_size,qemu_check_nic_model (&nd_table[0 ],"virtio" ); dev =qdev_create (NULL ,"syborg,virtio-net" ); dev->nd = &nd_table[0 ]; -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev);sysbus_mmio_map (s,0 ,0xc000c000 );sysbus_connect_irq (s,0 , pic[9 ]);
hw/sysbus.c +1 -1
@@ -143,7 +143,7 @@ DeviceState *sysbus_create_varargs(const char *name, dev =qdev_create (NULL , name); s =sysbus_from_qdev (dev); -qdev_init (dev); +qdev_init_nofail (dev);if (addr != (target_phys_addr_t)-1 ) {sysbus_mmio_map (s,0 , addr); }
hw/unin_pci.c +1 -1
@@ -253,7 +253,7 @@ PCIBus *pci_pmac_init(qemu_irq *pic)/* Use values found on a real PowerMac */ /* Uninorth main bus */ dev =qdev_create (NULL ,"Uni-north main" ); -qdev_init (dev); +qdev_init_nofail (dev); s =sysbus_from_qdev (dev); d =FROM_SYSBUS (UNINState, s); d->host_state.bus = pci_register_bus(NULL, "pci",
hw/usb-bus.c +1 -1
@@ -90,7 +90,7 @@ USBDevice *usb_create(USBBus *bus, const char *name)USBDevice *usb_create_simple (USBBus *bus,const char *name) {USBDevice *dev =usb_create (bus, name); -qdev_init (&dev->qdev); +qdev_init_nofail (&dev->qdev);return dev; }
hw/vga-pci.c +1 -1
@@ -140,7 +140,7 @@ int pci_vga_init(PCIBus *bus, dev =pci_create ("VGA" ,NULL );qdev_prop_set_uint32 (&dev->qdev,"bios-offset" , vga_bios_offset);qdev_prop_set_uint32 (&dev->qdev,"bios-size" , vga_bios_offset); -qdev_init (&dev->qdev); +qdev_init_nofail (&dev->qdev);return 0 ; }
hw/xilinx_ethlite.c +1 -1
@@ -214,7 +214,7 @@ xilinx_ethlite_create(NICInfo *nd, target_phys_addr_t base, qemu_irq irq, dev->nd = nd;qdev_prop_set_uint32 (dev,"txpingpong" , txpingpong);qdev_prop_set_uint32 (dev,"rxpingpong" , rxpingpong); -qdev_init (dev); +qdev_init_nofail (dev);sysbus_mmio_map (sysbus_from_qdev (dev),0 , base);sysbus_connect_irq (sysbus_from_qdev (dev),0 , irq);return dev;
hw/xilinx_intc.c +1 -1
@@ -153,7 +153,7 @@ xilinx_intc_create(target_phys_addr_t base, qemu_irq irq, int kind_of_intr) dev =qdev_create (NULL ,"xilinx,intc" );qdev_prop_set_uint32 (dev,"kind-of-intr" , kind_of_intr); -qdev_init (dev); +qdev_init_nofail (dev);sysbus_mmio_map (sysbus_from_qdev (dev),0 , base);sysbus_connect_irq (sysbus_from_qdev (dev),0 , irq);return dev;
hw/xilinx_timer.c +1 -1
@@ -198,7 +198,7 @@ xilinx_timer_create(target_phys_addr_t base, qemu_irq irq, int nr, int freq) dev =qdev_create (NULL ,"xilinx,timer" );qdev_prop_set_uint32 (dev,"nr-timers" , nr);qdev_prop_set_uint32 (dev,"frequency" , freq); -qdev_init (dev); +qdev_init_nofail (dev);sysbus_mmio_map (sysbus_from_qdev (dev),0 , base);sysbus_connect_irq (sysbus_from_qdev (dev),0 , irq);return dev;
usb-linux.c +2 -2
@@ -1387,7 +1387,7 @@ static int usb_host_auto_scan(void *opaque, int bus_num, int addr,/* We got a match */ -/* Allredy attached ? */ +/* Already attached ? */ if (hostdev_find (bus_num, addr))return 0 ;
@@ -1395,7 +1395,7 @@ static int usb_host_auto_scan(void *opaque, int bus_num, int addr, dev =usb_host_device_open_addr (bus_num, addr, product_name);if (dev) -qdev_init (&dev->qdev); +qdev_init_nofail (&dev->qdev); }return 0 ;
Signature · 1 line
1.6.2.5
PATCH 5/6 Make isa_create() terminate program on failure
In reply to Markus ArmbrusterCallers don't check the return value anyway.
--- hw/isa-bus.c | 8 +++----- 1 files changed, 3 insertions(+), 5 deletions(-)
hw/isa-bus.c +3 -5
@@ -114,8 +114,8 @@ ISADevice *isa_create(const char *name)DeviceState *dev;if (!isabus) { -fprintf (stderr,"Tried to create isa device %s with no isa bus present.\n" , name); -return NULL ; +hw_error ("Tried to create isa device %s with no isa bus present.\n" , + name); } dev =qdev_create (&isabus->qbus, name);return DO_UPCAST (ISADevice, qdev, dev);
@@ -126,9 +126,7 @@ ISADevice *isa_create_simple(const char *name)ISADevice *dev; dev =isa_create (name); -if (qdev_init (&dev->qdev) !=0 ) { -return NULL ; - } +qdev_init_nofail (&dev->qdev);return dev; }
Signature · 1 line
1.6.2.5
PATCH 6/6 Warn if value of qdev_init() isn't checked
In reply to Markus ArmbrusterAfter qdev_init() fails, the device is gone. Failure to check runs a high risk of use-after-free.
--- hw/qdev.h | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)
hw/qdev.h +1 -1
@@ -89,7 +89,7 @@ struct CompatProperty {DeviceState *qdev_create (BusState *bus,const char *name);DeviceState *qdev_device_add (QemuOpts *opts); -int qdev_init (DeviceState *dev); +int qdev_init (DeviceState *dev) __attribute__((warn_unused_result));void qdev_init_nofail (DeviceState *dev);void qdev_free (DeviceState *dev);
Signature · 1 line
1.6.2.5
Quoted 10 lines · Markus Armbruster
qdev_init()can fail since commit81a322d4. Most callers don't bother to check. This is less serious than it sounds, because they typically work with devices for whichqdev_init()never fails. It's still unclean.The last patch makes the compiler warn when the value of
qdev_init()isn't used. If that warning triggers after merging this series, more unchecked calls have crept in since the base of this series (commit2637c754). Feel free to ask me for a respin then.
Quoted 7 lines
Markus Armbruster (6):
Move function definitions out ofxilinx.h
Makeqdev_init()destroy the device on failure
Check return value ofqdev_init()
Newqdev_init_nofail()
Makeisa_create()terminate program on failure
Warn if value ofqdev_init()isn't checked
Whole series looks good to me, nice cleanup.
cheers,
Gerd