aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2015-09-21qapi: New QAPISchema intermediate reperesentationMarkus Armbruster
The QAPI code generators work with a syntax tree (nested dictionaries) plus a few symbol tables (also dictionaries) on the side. They have clearly outgrown these simple data structures. There's lots of rummaging around in dictionaries, and information is recomputed on the fly. For the work I'm going to do, I want more clearly defined and more convenient interfaces. Going forward, I also want less coupling between the back-ends and the syntax tree, to make messing with the syntax easier. Create a bunch of classes to represent QAPI schemata. Have the QAPISchema initializer call the parser, then walk the syntax tree to create the new internal representation, and finally perform semantic analysis. Shortcut: the semantic analysis still relies on existing check_exprs() to do the actual semantic checking. All this code needs to move into the classes. Mark as TODO. Simple unions are lowered to flat unions. Flat unions and structs are represented as a more general object type. Catching name collisions in generated code would be nice. Mark as TODO. We generate array types eagerly, even though most of them aren't used. Mark as TODO. Nothing uses the new intermediate representation just yet, thus no change to generated files. Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Daniel P. Berrange <berrange@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com>
2015-09-21qapi: Rename class QAPISchema to QAPISchemaParserMarkus Armbruster
I want to name a new class QAPISchema. While there, make it a new-style class. Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Daniel P. Berrange <berrange@redhat.com> Message-Id: <1442401589-24189-2-git-send-email-armbru@redhat.com>
2015-09-19Merge remote-tracking branch 'remotes/afaerber/tags/qom-devices-for-peter' ↵Peter Maydell
into staging QOM infrastructure fixes and device conversions * QOM API error handling fixes * Performance improvements for device GPIO property creation * Remaining conversion of QEMUMachine to QOM # gpg: Signature made Sat 19 Sep 2015 15:40:44 BST using RSA key ID 3E7E013F # gpg: Good signature from "Andreas Färber <afaerber@suse.de>" # gpg: aka "Andreas Färber <afaerber@suse.com>" * remotes/afaerber/tags/qom-devices-for-peter: (21 commits) machine: Eliminate QEMUMachine and qemu_register_machine() Revert use of DEFINE_MACHINE() for registrations of multiple machines Use DEFINE_MACHINE() to register all machines mac_world: Break long line machine: DEFINE_MACHINE() macro exynos4: Declare each QEMUMachine as a separate variable exynos4: Use MachineClass instead of exynos4_machines array exynos4: Use EXYNOS4210_NCPUS instead of max_cpus on error message machine: Set MachineClass::name automatically machine: Ensure all TYPE_MACHINE subclasses have the right suffix mac99: Use MACHINE_TYPE_NAME to encode class name s390: Rename s390-ccw-virtio-2.4 class name to use MACHINE_TYPE_NAME s390-virtio: Rename machine class name to use MACHINE_TYPE_NAME pseries: Rename machine class names to use MACHINE_TYPE_NAME arm: Rename virt machine class to use MACHINE_TYPE_NAME vexpress: Rename machine classes to use MACHINE_TYPE_NAME vexpress: Don't set name on abstract class machine: MACHINE_TYPE_NAME macro qdev: Do not use slow [*] expansion for GPIO creation qom: Fix invalid error check in property_get_str() ... Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2015-09-19machine: Eliminate QEMUMachine and qemu_register_machine()Eduardo Habkost
The struct is not used anymore and can be eliminated. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19Revert use of DEFINE_MACHINE() for registrations of multiple machinesAndreas Färber
The script used for converting from QEMUMachine had used one DEFINE_MACHINE() per machine registered. In cases where multiple machines are registered from one source file, avoid the excessive generation of module init functions by reverting this unrolling. Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19Use DEFINE_MACHINE() to register all machinesEduardo Habkost
Convert all machines to use DEFINE_MACHINE() instead of QEMUMachine automatically using a script. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> [AF: Style cleanups, convert imx25_pdk machine] Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19mac_world: Break long lineEduardo Habkost
Coding style change only. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19machine: DEFINE_MACHINE() macroEduardo Habkost
The macro will allow easy registration of a TYPE_MACHINE subclass, using only the machine name and a MachineClass initialization function as parameter. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19exynos4: Declare each QEMUMachine as a separate variableEduardo Habkost
This will make the code follow the same pattern used for other machines, and will make it easier to automatically convert the code to be QOM-based. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19exynos4: Use MachineClass instead of exynos4_machines arrayEduardo Habkost
We don't need a QEMUMachine array to query max_cpus, if we can get the corresponding MachineClass. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19exynos4: Use EXYNOS4210_NCPUS instead of max_cpus on error messageEduardo Habkost
The code is checking smp_cpus against EXYNOS4210_NCPUS, not against max_cpus, so use EXYNOS4210_NCPUS in the error message for consistency. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19machine: Set MachineClass::name automaticallyEduardo Habkost
Now all TYPE_MACHINE subclasses use MACHINE_TYPE_NAME to generate the class name. So instead of requiring each subclass to set MachineClass::name manually, we can now set it automatically at the TYPE_MACHINE class_base_init() function. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Reviewed-by: Marcel Apfelbaum <marcel@redhat.com> [AF/ehabkost: Updated for s390-ccw machines] [AF: Cleanup of intermediate virt and vexpress name handling] Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19machine: Ensure all TYPE_MACHINE subclasses have the right suffixEduardo Habkost
Now that all non-abstract TYPE_MACHINE subclasses have the -machine suffix, add an assert to ensure this will be always true. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Reviewed-by: Marcel Apfelbaum <marcel@redhat.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19mac99: Use MACHINE_TYPE_NAME to encode class nameEduardo Habkost
It will result in exactly the same class name, but it will make the code consistent with the other classes. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19s390: Rename s390-ccw-virtio-2.4 class name to use MACHINE_TYPE_NAMEEduardo Habkost
Machine class names should use the "-machine" suffix to allow class-name-based machine class lookup to work. Rename the s390-ccw-virtio-2.4 machine class using the MACHINE_TYPE_NAME macro. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> [AF/ehabkost: Updated for 2.5 machine] Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19s390-virtio: Rename machine class name to use MACHINE_TYPE_NAMEEduardo Habkost
Machine class names should use the "-machine" suffix to allow class-name-based machine class lookup to work. Rename the s390-virtio machine class using the MACHINE_TYPE_NAME macro. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Acked-by: Cornelia Huck <cornelia.huck@de.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19pseries: Rename machine class names to use MACHINE_TYPE_NAMEEduardo Habkost
Machine class names should use the "-machine" suffix to allow class-name-based machine class lookup to work. Rename the the pseries machine classes using the MACHINE_TYPE_NAME macro. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19arm: Rename virt machine class to use MACHINE_TYPE_NAMEEduardo Habkost
Machine class names should use the "-machine" suffix to allow class-name-based machine class lookup to work. Rename the arm virt machine class using the MACHINE_TYPE_NAME macro. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19vexpress: Rename machine classes to use MACHINE_TYPE_NAMEEduardo Habkost
Machine class names should use the "-machine" suffix to allow class-name-based machine class lookup to work. Rename the vexpress machine classes using the MACHINE_TYPE_NAME macro. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> [AF: Introduce VEXPRESS_*_MACHINE_NAME] Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19vexpress: Don't set name on abstract classEduardo Habkost
The MachineClass::name field won't be ever be used on TYPE_VEXPRESS, as it is an abstract class and the machine class lookup code explicitly skips abstract classes. We can remove it to make the code simpler. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19machine: MACHINE_TYPE_NAME macroEduardo Habkost
The macro will be useful to ensure the machine class names follow the right format to make machine class lookup by class name work correctly. Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Reviewed-by: Marcel Apfelbaum <marcel@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19qdev: Do not use slow [*] expansion for GPIO creationPavel Fedin
Expansion of [*] suffix is very slow because index expansion is done using trial and error strategy, starting every time from zero and retrying with the next index until insertion succeeds. With large number of already added properties this process takes huge amount of time (O(n^2) complexity). Some architectures (like ARM) use very large amount of IRQ pins in interrupt controller models. This flaw makes machine startup extremely slow (~20 seconds for ARM64 with 32 CPUs). This patch decreases this time down to ~10 seconds. Also in qdev_init_gpio_out_named() memset() is now called only once for the whole array instead of per-cell cleaning Signed-off-by: Pavel Fedin <p.fedin@samsung.com> Reviewed-by: Daniel P. Berrange <berrange@redhat.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19qom: Fix invalid error check in property_get_str()Markus Armbruster
When a function returns a null pointer on error and only on error, you can do if (!foo(foos, errp)) { ... handle error ... } instead of the more cumbersome Error *err = NULL; if (!foo(foos, &err)) { error_propagate(errp, err); ... handle error ... } A StringProperty's getter, however, may return null on success! We then fail to call visit_type_str(). Screwed up in 6a146eb, v1.1. Fails tests/qom-test in my current, heavily hacked QAPI branch. No reproducer for master known (but I didn't look hard). Cc: Anthony Liguori <anthony@codemonkey.ws> Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Cc: qemu-stable@nongnu.org Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-19qom: Do not reuse errp after a possible errorMarkus Armbruster
The argument for an Error **errp parameter must point to a null pointer. If it doesn't, and an error happens, error_set() fails its assertion. Instead of foo(foos, errp); bar(bars, errp); you need to do something like Error *err = NULL; foo(foos, &err); if (err) { error_propagate(errp, err); goto out; } bar(bars, errp); out: Screwed up in commit 0e55884 (v1.3.0): property_get_bool(). Screwed up in commit 1f21772 (v2.1.0): object_property_get_enum() and object_property_get_uint16List(). Screwed up in commit a8e3fbe (v2.4.0): property_get_enum(), property_set_enum(). Found by inspection, no actual crashes observed. Fix them up. Cc: Anthony Liguori <anthony@codemonkey.ws> Cc: Hu Tao <hutao@cn.fujitsu.com> Cc: Daniel P. Berrange <berrange@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Daniel P. Berrange <berrange@redhat.com> Cc: qemu-stable@nongnu.org Signed-off-by: Andreas Färber <afaerber@suse.de>
2015-09-18cocoa: Suppress Cocoa window with -displayRainer Müller
Do not open a Cocoa window when another display is selected that will be initialized later. The Cocoa display cannot be selected with -display, so there is no need to check its argument. Signed-off-by: Rainer Müller <raimue@codingfarm.de> Reviewed-by: Andreas Färber <andreas.faerber@web.de> Message-id: 1441807710-25431-1-git-send-email-raimue@codingfarm.de Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2015-09-18Merge remote-tracking branch 'remotes/jnsnow/tags/ide-pull-request' into stagingPeter Maydell
# gpg: Signature made Fri 18 Sep 2015 15:59:02 BST using RSA key ID AAFC390E # gpg: Good signature from "John Snow (John Huston) <jsnow@redhat.com>" * remotes/jnsnow/tags/ide-pull-request: ahci: clean up initial d2h semantics ahci: remove cmd_fis argument from write_fis_d2h ahci: fix signature generation ahci: remove dead reset code atapi: abort transfers with 0 byte limits ide: fix ATAPI command permissions ide-test: add cdrom dma test ide-test: add cdrom pio test qtest/ahci: export generate_pattern qtest/ahci: use generate_pattern everywhere ide: unify io_buffer_offset increments Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2015-09-18ahci: clean up initial d2h semanticsJohn Snow
with write_fis_d2h and signature generation tidied up, let's adjust the initial d2h semantics to make more sense. The initial d2h is considered delivered if there is guest memory to save it to. Signed-off-by: John Snow <jsnow@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Message-id: 1441140641-17631-5-git-send-email-jsnow@redhat.com
2015-09-18ahci: remove cmd_fis argument from write_fis_d2hJohn Snow
It's no longer used. We used to generate a D2H FIS based upon the command FIS that prompted the update, but in reality, the D2H FIS is generated purely from register state. cmd_fis is vestigial, so get rid of it. Signed-off-by: John Snow <jsnow@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Message-id: 1441140641-17631-4-git-send-email-jsnow@redhat.com
2015-09-18ahci: fix signature generationJohn Snow
The initial register device-to-host FIS no longer needs to specially set certain fields, as these can be handled generically by setting those fields explicitly with the signatures we want at port reset time. (1) Signatures are decomposed into their four component registers and set upon (AHCI) port reset. (2) the signature cache register is no longer set manually per-each device type, but instead just once during ahci_init_d2h. Signed-off-by: John Snow <jsnow@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Message-id: 1441140641-17631-3-git-send-email-jsnow@redhat.com
2015-09-18ahci: remove dead reset codeJohn Snow
This check is dead due to an earlier conditional. AHCI does not currently support hotplugging, so checks to see if devices are present or not are useless. Remove it. Reported-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: John Snow <jsnow@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Message-id: 1441140641-17631-2-git-send-email-jsnow@redhat.com
2015-09-18atapi: abort transfers with 0 byte limitsJohn Snow
We're supposed to abort on transfers like this, unless we fill Word 125 of our IDENTIFY data with a default transfer size, which we don't currently do. This is an ATA error, not a SCSI/ATAPI one. See ATA8-ACS3 sections 7.17.6.49 or 7.21.5. If we don't do this, QEMU will loop forever trying to transfer zero bytes, which isn't particularly useful. Signed-off-by: John Snow <jsnow@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Message-id: 1442253685-23349-2-git-send-email-jsnow@redhat.com
2015-09-18ide: fix ATAPI command permissionsJohn Snow
We're a little too lenient with what we'll let an ATAPI drive handle. Clamp down on the IDE command execution table to remove CD_OK permissions from commands that are not and have never been ATAPI commands. For ATAPI command validity, please see: - ATA4 Section 6.5 ("PACKET Command feature set") - ATA8/ACS Section 4.3 ("The PACKET feature set") - ACS3 Section 4.3 ("The PACKET feature set") ACS3 has a historical command validity table in Table B.4 ("Historical Command Assignments") that can be referenced to find when a command was introduced, deprecated, obsoleted, etc. The only reference for ATAPI command validity is by checking that version's PACKET feature set section. ATAPI was introduced by T13 into ATA4, all commands retired prior to ATA4 therefore are assumed to have never been ATAPI commands. Mandatory commands, as listed in ATA8-ACS3, are: - DEVICE RESET - EXECUTE DEVICE DIAGNOSTIC - IDENTIFY DEVICE - IDENTIFY PACKET DEVICE - NOP - PACKET - READ SECTOR(S) - SET FEATURES Optional commands as listed in ATA8-ACS3, are: - FLUSH CACHE - READ LOG DMA EXT - READ LOG EXT - WRITE LOG DMA EXT - WRITE LOG EXT All other commands are illegal to send to an ATAPI device and should be rejected by the device. CD_OK removal justifications: 0x06 WIN_DSM Defined in ACS2. Not valid for ATAPI. 0x21 WIN_READ_ONCE Retired in ATA5. Not ATAPI in ATA4. 0x94 WIN_STANDBYNOW2 Retired in ATA4. Did not coexist with ATAPI. 0x95 WIN_IDLEIMMEDIATE2 Retired in ATA4. Did not coexist with ATAPI. 0x96 WIN_STANDBY2 Retired in ATA4. Did not coexist with ATAPI. 0x97 WIN_SETIDLE2 Retired in ATA4. Did not coexist with ATAPI. 0x98 WIN_CHECKPOWERMODE2 Retired in ATA4. Did not coexist with ATAPI. 0x99 WIN_SLEEPNOW2 Retired in ATA4. Did not coexist with ATAPI. 0xE0 WIN_STANDBYNOW1 Not part of ATAPI in ATA4, ACS or ACS3. 0xE1 WIN_IDLEIMMDIATE Not part of ATAPI in ATA4, ACS or ACS3. 0xE2 WIN_STANDBY Not part of ATAPI in ATA4, ACS or ACS3. 0xE3 WIN_SETIDLE1 Not part of ATAPI in ATA4, ACS or ACS3. 0xE4 WIN_CHECKPOWERMODE1 Not part of ATAPI in ATA4, ACS or ACS3. 0xE5 WIN_SLEEPNOW1 Not part of ATAPI in ATA4, ACS or ACS3. 0xF8 WIN_READ_NATIVE_MAX Obsoleted in ACS3. Not ATAPI in ATA4 or ACS. This patch fixes a divide by zero fault that can be caused by sending the WIN_READ_NATIVE_MAX command to an ATAPI drive, which causes it to attempt to use zeroed CHS values to perform sector arithmetic. Reported-by: Qinghao Tang <luodalongde@gmail.com> Signed-off-by: John Snow <jsnow@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Message-id: 1441816082-21031-1-git-send-email-jsnow@redhat.com CC: qemu-stable@nongnu.org
2015-09-18ide-test: add cdrom dma testJohn Snow
Now, test the DMA functionality of the ATAPI drive. Signed-off-by: John Snow <jsnow@redhat.com> Message-id: 1441926555-19471-5-git-send-email-jsnow@redhat.com
2015-09-18ide-test: add cdrom pio testJohn Snow
Add a simple read test for ATAPI devices, using the PIO mechanism. Signed-off-by: John Snow <jsnow@redhat.com> Message-id: 1441926555-19471-4-git-send-email-jsnow@redhat.com
2015-09-18qtest/ahci: export generate_patternJohn Snow
Share the pattern function for ide and ahci test. Signed-off-by: John Snow <jsnow@redhat.com> Message-id: 1441926555-19471-3-git-send-email-jsnow@redhat.com
2015-09-18qtest/ahci: use generate_pattern everywhereJohn Snow
Fix the pattern generation to actually be interesting, and make sure all buffers in the ahci-test actually use it. Signed-off-by: John Snow <jsnow@redhat.com> Message-id: 1441926555-19471-2-git-send-email-jsnow@redhat.com
2015-09-18Merge remote-tracking branch 'remotes/armbru/tags/pull-error-2015-09-18' ↵Peter Maydell
into staging Error reporting patches # gpg: Signature made Fri 18 Sep 2015 13:42:49 BST using RSA key ID EB918653 # gpg: Good signature from "Markus Armbruster <armbru@redhat.com>" # gpg: aka "Markus Armbruster <armbru@pond.sub.org>" * remotes/armbru/tags/pull-error-2015-09-18: memory: Fix bad error handling in memory_region_init_ram_ptr() loader: Fix memory_region_init_resizeable_ram() error handling Fix bad error handling after memory_region_init_ram() error: New error_fatal MAINTAINERS: Add "Error reporting" entry error: Copy location information in error_copy() hmp: Allow for error message hints on HMP error: only prepend timestamp on stderr Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2015-09-18memory: Fix bad error handling in memory_region_init_ram_ptr()Markus Armbruster
Commit ef701d7 screwed up handling of out-of-memory conditions. Before the commit, we report the error and exit(1), in one place. The commit lifts the error handling up the call chain some, to three places. Fine. Except it uses &error_abort in these places, changing the behavior from exit(1) to abort(), and thus undoing the work of commit 3922825 "exec: Don't abort when we can't allocate guest memory". The previous two commits fixed one of the three places, another one was fixed in commit 33e0eb5. This commit fixes the third one. Signed-off-by: Markus Armbruster <armbru@redhat.com> Message-Id: <1441983105-26376-5-git-send-email-armbru@redhat.com> Reviewed-by: Peter Crosthwaite <crosthwaite.peter@gmail.com>
2015-09-18loader: Fix memory_region_init_resizeable_ram() error handlingMarkus Armbruster
Commit ef701d7 screwed up handling of out-of-memory conditions. Before the commit, we report the error and exit(1), in one place. The commit lifts the error handling up the call chain some, to three places. Fine. Except it uses &error_abort in these places, changing the behavior from exit(1) to abort(), and thus undoing the work of commit 3922825 "exec: Don't abort when we can't allocate guest memory". The previous commit fixed up uses of memory_region_init_ram(). One of them was replaced by memory_region_init_resizeable_ram() [sic!] in commit a166614, so Coccinelle missed it. Fix it up. Signed-off-by: Markus Armbruster <armbru@redhat.com> Message-Id: <1441983105-26376-4-git-send-email-armbru@redhat.com> Reviewed-by: Peter Crosthwaite <crosthwaite.peter@gmail.com>
2015-09-18Fix bad error handling after memory_region_init_ram()Markus Armbruster
Symptom: $ qemu-system-x86_64 -m 10000000 Unexpected error in ram_block_add() at /work/armbru/qemu/exec.c:1456: upstream-qemu: cannot set up guest memory 'pc.ram': Cannot allocate memory Aborted (core dumped) Root cause: commit ef701d7 screwed up handling of out-of-memory conditions. Before the commit, we report the error and exit(1), in one place, ram_block_add(). The commit lifts the error handling up the call chain some, to three places. Fine. Except it uses &error_abort in these places, changing the behavior from exit(1) to abort(), and thus undoing the work of commit 3922825 "exec: Don't abort when we can't allocate guest memory". The three places are: * memory_region_init_ram() Commit 4994653 (right after commit ef701d7) lifted the error handling further, through memory_region_init_ram(), multiplying the incorrect use of &error_abort. Later on, imitation of existing (bad) code may have created more. * memory_region_init_ram_ptr() The &error_abort is still there. * memory_region_init_rom_device() Doesn't need fixing, because commit 33e0eb5 (soon after commit ef701d7) lifted the error handling further, and in the process changed it from &error_abort to passing it up the call chain. Correct, because the callers are realize() methods. Fix the error handling after memory_region_init_ram() with a Coccinelle semantic patch: @r@ expression mr, owner, name, size, err; position p; @@ memory_region_init_ram(mr, owner, name, size, ( - &error_abort + &error_fatal | err@p ) ); @script:python@ p << r.p; @@ print "%s:%s:%s" % (p[0].file, p[0].line, p[0].column) When the last argument is &error_abort, it gets replaced by &error_fatal. This is the fix. If the last argument is anything else, its position is reported. This lets us check the fix is complete. Four positions get reported: * ram_backend_memory_alloc() Error is passed up the call chain, ultimately through user_creatable_complete(). As far as I can tell, it's callers all handle the error sanely. * fsl_imx25_realize(), fsl_imx31_realize(), dp8393x_realize() DeviceClass.realize() methods, errors handled sanely further up the call chain. We're good. Test case again behaves: $ qemu-system-x86_64 -m 10000000 qemu-system-x86_64: cannot set up guest memory 'pc.ram': Cannot allocate memory [Exit 1 ] The next commits will repair the rest of commit ef701d7's damage. Signed-off-by: Markus Armbruster <armbru@redhat.com> Message-Id: <1441983105-26376-3-git-send-email-armbru@redhat.com> Reviewed-by: Peter Crosthwaite <crosthwaite.peter@gmail.com>
2015-09-18error: New error_fatalMarkus Armbruster
Similar to error_abort, but doesn't report where the error was created, and terminates the process with exit(1) rather than abort(). Signed-off-by: Markus Armbruster <armbru@redhat.com> Message-Id: <1441983105-26376-2-git-send-email-armbru@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Peter Crosthwaite <crosthwaite.peter@gmail.com>
2015-09-18MAINTAINERS: Add "Error reporting" entryMarkus Armbruster
Error reporting work has been flowing through my tree for a while. Time for MAINTAINERS to catch up. Signed-off-by: Markus Armbruster <armbru@redhat.com> Message-Id: <1442057396-21989-1-git-send-email-armbru@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Peter Crosthwaite <crosthwaite.peter@gmail.com>
2015-09-18error: Copy location information in error_copy()Eric Blake
Commit 1e9b65bb forgot to propagate source information to copied errors. Signed-off-by: Eric Blake <eblake@redhat.com> Message-Id: <1441902890-23064-1-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-09-18hmp: Allow for error message hints on HMPEric Blake
Commits 7216ae3d and d2828429 disabled some error message hints, all because a change to use modern error reporting meant that the hint would be output prior to the actual error. Fix this by making hints a first-class member of Error. For example, we are now back to the pleasant: $ qemu-system-x86_64 --nodefaults -S --vnc :0 --chardev null,id=, qemu-system-x86_64: --chardev null,id=,: Parameter 'id' expects an identifier Identifiers consist of letters, digits, '-', '.', '_', starting with a letter. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Message-Id: <1441901956-21991-1-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-09-18error: only prepend timestamp on stderrStefan Hajnoczi
The -msg timestamp=on option prepends a timestamp to error messages. This is useful on stderr where it allows users to identify when an error was raised. Timestamps do not make sense on the monitor since error_report() is called in response to a synchronous monitor command and the user already knows "when" the command was issued. Additionally, the rest of the monitor conversation lacks timestamps so the error timestamp cannot be correlated with other activity. Only prepend timestamps on stderr. This fixes libvirt's 'drive_del' processing, which did not expect a timestamp. Other QEMU monitor clients are probably equally confused by timestamps on monitor error messages. Cc: Markus Armbruster <armbru@redhat.com> Cc: Seiji Aguchi <seiji.aguchi@hds.com> Cc: Frank Schreuder <fschreuder@transip.nl> Cc: Daniel P. Berrange <berrange@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Message-Id: <1439212541-16997-1-git-send-email-stefanha@redhat.com> Reviewed-by: Daniel P. Berrange <berrange@redhat.com> Tested-by: Frank Schreuder <fschreuder@transip.nl> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-09-18Merge remote-tracking branch 'remotes/lalrae/tags/mips-20150918' into stagingPeter Maydell
MIPS patches 2015-09-18 Changes: * fixes for rdhwr, tlbwr, mtc0, recip.fmt, rsqrt.fmt and daui instructions * removal of MIPS_DEBUG code * use tcg_gen_extrh_i64_i32() * improve random tlb index generation in cpu_mips_get_random() * exception handling improvements to correctly restore icount # gpg: Signature made Fri 18 Sep 2015 12:15:28 BST using RSA key ID 0B29DA6B # gpg: Good signature from "Leon Alrae <leon.alrae@imgtec.com>" * remotes/lalrae/tags/mips-20150918: target-mips: improve exception handling target-mips: correct MTC0 instruction on MIPS64 target-mips: add missing restriction in DAUI instruction target-mips: fix corner case in TLBWR causing QEMU to hang pic32: use LCG algorithm for generated random index of TLBWR instruction target-mips: get rid of MIPS_DEBUG_SIGN_EXTENSIONS target-mips: get rid of MIPS_DEBUG target-mips: Fix RDHWR on CP0.Count target-mips: remove wrong checks for recip.fmt and rsqrt.fmt target-mips: Use tcg_gen_extrh_i64_i32 Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2015-09-18target-mips: improve exception handlingPavel Dovgaluk
This patch improves exception handling in MIPS. Instructions generate several types of exceptions. When exception is generated, it breaks the execution of the current translation block. Implementation of the exceptions handling does not correctly restore icount for the instruction which caused the exception. In most cases icount will be decreased by the value equal to the size of TB. This patch passes pointer to the translation block internals to the exception handler. It allows correct restoring of the icount value. Signed-off-by: Pavel Dovgalyuk <pavel.dovgaluk@ispras.ru> Reviewed-by: Aurelien Jarno <aurelien@aurel32.net> Reviewed-by: Leon Alrae <leon.alrae@imgtec.com> [leon.alrae@imgtec.com: avoid retranslation in linux-user SC, break lines which are over 80 chars, remove v3 changelog from the commit message] Signed-off-by: Leon Alrae <leon.alrae@imgtec.com>
2015-09-18target-mips: correct MTC0 instruction on MIPS64Leon Alrae
MTC0 on a 64-bit processor should move entire 64-bit GPR content to CP0 register. Signed-off-by: Leon Alrae <leon.alrae@imgtec.com> Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
2015-09-18target-mips: add missing restriction in DAUI instructionLeon Alrae
rs cannot be the zero register, Reserved Instruction exception must be signalled for this case. Signed-off-by: Leon Alrae <leon.alrae@imgtec.com> Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
2015-09-18target-mips: fix corner case in TLBWR causing QEMU to hangLeon Alrae
cpu_mips_get_random() function is used to generate a random index from CP0.Wired to TLBSize-1 range. Current implementation avoids generating the same as before value, hence the while loop. If the guest sets CP0.Wired to TLBSize-1 (which actually does not sound to be very practical) QEMU will get stuck in the loop infinitely as we always generate the same index. Signed-off-by: Leon Alrae <leon.alrae@imgtec.com> Reviewed-by: Aurelien Jarno <aurelien@aurel32.net>
generated by cgit v1.2.3 (git 2.26.2) at 2024-12-21 02:23:29 +0000