tpm: Fix error reporting, improve help
cover tpm: Fix error reporting, improve help
The alternative to PATCH 1 is a proper conversion to Error, as discussed in
Subject: Re: What is TYPE_TPM_TIS_ISA? (Not an ISA Device)
Message-ID: <[email protected]>
Such a conversion would be much too invasive for 5.1. Going the other way, like PATCH 1 does, is simple enough to be considered for 5.1.
To make this series a complete alternative to Philippe's "[PATCH-for-5.1 v2 0/2] tpm: Improve error reporting", PATCH 2 improves help.
Markus Armbruster (2):
Revert "tpm: Clean up error reporting in tpm_init_tpmdev()"
tpm: Improve help on TPM types when none are available
include/sysemu/tpm.h | 2 +- softmmu/vl.c | 4 +++- stubs/tpm.c | 3 ++- tpm.c | 43 ++++++++++++++++++++++++++++++------------- 4 files changed, 36 insertions(+), 16 deletions(-)
Signature · 1 line
2.26.2
PATCH 1/2 Revert "tpm: Clean up error reporting in tpm_init_tpmdev()"
This reverts commit d10e05f15d5c3dd5e5cc59c5dfff460d89d48580.
We report some -tpmdev failures, but then continue as if all was fine. Reproducer:
$ qemu-system-x86_64 -nodefaults -S -display none -monitor stdio -chardev null,id=tpm0 -tpmdev emulator,id=tpm0,chardev=chrtpm -device tpm-tis,tpmdev=tpm0
qemu-system-x86_64: -tpmdev emulator,id=tpm0,chardev=chrtpm: tpm-emulator: tpm chardev 'chrtpm' not found.
qemu-system-x86_64: -tpmdev emulator,id=tpm0,chardev=chrtpm: tpm-emulator: Could not cleanly shutdown the TPM: No such file or directory
QEMU 5.0.90 monitor - type 'help' for more information
(qemu) qemu-system-x86_64: -device tpm-tis,tpmdev=tpm0: Property 'tpm-tis.tpmdev' can't find value 'tpm0'
$ echo $?
1This is a regression caused by commit d10e05f15d "tpm: Clean up error reporting in tpm_init_tpmdev()". It's incomplete: be->create(opts) continues to use error_report(), and we don't set an error when it fails.
I figure converting the create() methods to Error would make some sense, but I'm not sure it's worth the effort right now. Revert the broken commit instead, and add a comment to tpm_init_tpmdev().
Straightforward conflict in tpm.c resolved.
--- include/sysemu/tpm.h | 2 +- softmmu/vl.c | 4 +++- stubs/tpm.c | 3 ++- tpm.c | 30 +++++++++++++++++++++--------- 4 files changed, 27 insertions(+), 12 deletions(-)
include/sysemu/tpm.h +1 -1
@@ -16,7 +16,7 @@#include "qom/object.h" int tpm_config_parse (QemuOptsList *opts_list,const char *optarg); -void tpm_init (void ); +int tpm_init (void );void tpm_cleanup (void ); typedef enum TPMVersion {
softmmu/vl.c +3 -1
@@ -4262,7 +4262,9 @@ void qemu_init(int argc, char **argv, char **envp) user_creatable_add_opts_foreach, object_create_delayed, &error_fatal); -tpm_init (); +if (tpm_init () <0 ) { +exit (1 ); + }blk_mig_init ();ram_mig_init ();
stubs/tpm.c +2 -1
@@ -10,8 +10,9 @@#include "sysemu/tpm.h" #include "hw/acpi/tpm.h" -void tpm_init (void ) +int tpm_init (void ) { +return 0 ; }void tpm_cleanup (void )
tpm.c +21 -9
@@ -81,26 +81,33 @@ TPMBackend *qemu_find_tpm_be(const char *id)static int tpm_init_tpmdev (void *dummy,QemuOpts *opts,Error **errp) { +/* +* Use of error_report() in a function with an Error ** parameter +* is suspicious. It is okay here. The parameter only exists to +* make the function usable with qemu_opts_foreach(). It is not +* actually used. +*/ const char *value;const char *id;const TPMBackendClass *be;TPMBackend *drv; +Error *local_err =NULL ;int i;if (!QLIST_EMPTY (&tpm_backends)) { -error_setg (errp,"Only one TPM is allowed." ); +error_report ("Only one TPM is allowed." );return 1 ; } id =qemu_opts_id (opts);if (id ==NULL ) { -error_setg (errp,QERR_MISSING_PARAMETER ,"id" ); +error_report (QERR_MISSING_PARAMETER ,"id" );return 1 ; } value =qemu_opt_get (opts,"type" );if (!value) { -error_setg (errp,QERR_MISSING_PARAMETER ,"type" ); +error_report (QERR_MISSING_PARAMETER ,"type" );tpm_display_backend_drivers ();return 1 ; }
@@ -108,14 +115,15 @@ static int tpm_init_tpmdev(void *dummy, QemuOpts *opts, Error **errp) i =qapi_enum_parse (&TpmType_lookup, value,-1 ,NULL ); be = i >=0 ?tpm_be_find_by_type (i) :NULL ;if (be ==NULL ) { -error_setg (errp,QERR_INVALID_PARAMETER_VALUE ,"type" , -"a TPM backend type" ); +error_report (QERR_INVALID_PARAMETER_VALUE , +"type" ,"a TPM backend type" );tpm_display_backend_drivers ();return 1 ; }/* validate backend specific opts */ -if (!qemu_opts_validate (opts, be->opts, errp)) { +if (!qemu_opts_validate (opts, be->opts, &local_err)) { +error_report_err (local_err);return 1 ; }
@@ -148,10 +156,14 @@ void tpm_cleanup(void)* Initialize the TPM. Process the tpmdev command line options describing the * TPM backend. */ -void tpm_init (void ) +int tpm_init (void ) { -qemu_opts_foreach (qemu_find_opts ("tpmdev" ), - tpm_init_tpmdev,NULL , &error_fatal); +if (qemu_opts_foreach (qemu_find_opts ("tpmdev" ), + tpm_init_tpmdev,NULL ,NULL )) { +return -1 ; + } + +return 0 ; }/*
Signature · 1 line
2.26.2
Re: [PATCH for-5.1 1/2] Revert "tpm: Clean up error reporting in tpm_init_tpmdev()"
Quoted 18 lines · Markus Armbruster
This reverts commit
d10e05f15d5c3dd5e5cc59c5dfff460d89d48580.We report some -tpmdev failures, but then continue as if all was fine. Reproducer:
$ qemu-system-x86_64 -nodefaults -S -display none -monitor stdio -chardev null,id=tpm0 -tpmdev emulator,id=tpm0,chardev=chrtpm -device tpm-tis,tpmdev=tpm0 qemu-system-x86_64: -tpmdev emulator,id=tpm0,chardev=chrtpm: tpm-emulator: tpm chardev 'chrtpm' not found. qemu-system-x86_64: -tpmdev emulator,id=tpm0,chardev=chrtpm: tpm-emulator: Could not cleanly shutdown the TPM: No such file or directory QEMU 5.0.90 monitor - type 'help' for more information (qemu) qemu-system-x86_64: -device tpm-tis,tpmdev=tpm0: Property 'tpm-tis.tpmdev' can't find value 'tpm0' $ echo $? 1This is a regression caused by commit
d10e05f15d"tpm: Clean up error reporting intpm_init_tpmdev()". It's incomplete:be->create(opts)continues to useerror_report(), and we don't set an error when it fails.
Quoted 7 lines
I figure converting the
create()methods to Error would make some sense, but I'm not sure it's worth the effort right now. Revert the broken commit instead, and add a comment totpm_init_tpmdev().Straightforward conflict in
tpm.cresolved.Signed-off-by Markus Armbruster <[email protected]>
Quoted 129 lines
--- include/sysemu/tpm.h | 2 +- softmmu/vl.c | 4 +++- stubs/tpm.c | 3 ++- tpm.c | 30 +++++++++++++++++++++--------- 4 files changed, 27 insertions(+), 12 deletions(-)include/sysemu/tpm.h +1
-1@@ -16,7 +16,7 @@#include "qom/object.h" int tpm_config_parse (QemuOptsList *opts_list,const char *optarg); -void tpm_init (void ); +int tpm_init (void );void tpm_cleanup (void ); typedef enum TPMVersion {softmmu/vl.c +3
-1@@ -4262,7 +4262,9 @@ void qemu_init(int argc, char **argv, char **envp) user_creatable_add_opts_foreach, object_create_delayed, &error_fatal); -tpm_init (); +if (tpm_init () <0 ) { +exit (1 ); + }blk_mig_init ();ram_mig_init ();stubs/tpm.c +2
-1@@ -10,8 +10,9 @@#include "sysemu/tpm.h" #include "hw/acpi/tpm.h" -void tpm_init (void ) +int tpm_init (void ) { +return 0 ; }void tpm_cleanup (void )tpm.c +21
-9@@ -81,26 +81,33 @@ TPMBackend *qemu_find_tpm_be(const char *id)static int tpm_init_tpmdev (void *dummy,QemuOpts *opts,Error **errp) { +/* +* Use of error_report() in a function with an Error ** parameter +* is suspicious. It is okay here. The parameter only exists to +* make the function usable with qemu_opts_foreach(). It is not +* actually used. +*/ const char *value;const char *id;const TPMBackendClass *be;TPMBackend *drv; +Error *local_err =NULL ;int i;if (!QLIST_EMPTY (&tpm_backends)) { -error_setg (errp,"Only one TPM is allowed." ); +error_report ("Only one TPM is allowed." );return 1 ; } id =qemu_opts_id (opts);if (id ==NULL ) { -error_setg (errp,QERR_MISSING_PARAMETER ,"id" ); +error_report (QERR_MISSING_PARAMETER ,"id" );return 1 ; } value =qemu_opt_get (opts,"type" );if (!value) { -error_setg (errp,QERR_MISSING_PARAMETER ,"type" ); +error_report (QERR_MISSING_PARAMETER ,"type" );tpm_display_backend_drivers ();return 1 ; }@@ -108,14 +115,15 @@ static int tpm_init_tpmdev(void *dummy, QemuOpts *opts, Error **errp) i =qapi_enum_parse (&TpmType_lookup, value,-1 ,NULL ); be = i >=0 ?tpm_be_find_by_type (i) :NULL ;if (be ==NULL ) { -error_setg (errp,QERR_INVALID_PARAMETER_VALUE ,"type" , -"a TPM backend type" ); +error_report (QERR_INVALID_PARAMETER_VALUE , +"type" ,"a TPM backend type" );tpm_display_backend_drivers ();return 1 ; }/* validate backend specific opts */ -if (!qemu_opts_validate (opts, be->opts, errp)) { +if (!qemu_opts_validate (opts, be->opts, &local_err)) { +error_report_err (local_err);return 1 ; }@@ -148,10 +156,14 @@ void tpm_cleanup(void)* Initialize the TPM. Process the tpmdev command line options describing the * TPM backend. */ -void tpm_init (void ) +int tpm_init (void ) { -qemu_opts_foreach (qemu_find_opts ("tpmdev" ), - tpm_init_tpmdev,NULL , &error_fatal); +if (qemu_opts_foreach (qemu_find_opts ("tpmdev" ), + tpm_init_tpmdev,NULL ,NULL )) { +return -1 ; + } + +return 0 ; }/*
Stefan Berger
Reviewed-byRe: [PATCH for-5.1 1/2] Revert "tpm: Clean up error reporting in tpm_init_tpmdev()"
Quoted 19 lines · Philippe Mathieu-Daudé
Quoted 18 lines · Markus Armbruster
This reverts commit
d10e05f15d5c3dd5e5cc59c5dfff460d89d48580.We report some -tpmdev failures, but then continue as if all was fine. Reproducer:
$ qemu-system-x86_64 -nodefaults -S -display none -monitor stdio -chardev null,id=tpm0 -tpmdev emulator,id=tpm0,chardev=chrtpm -device tpm-tis,tpmdev=tpm0 qemu-system-x86_64: -tpmdev emulator,id=tpm0,chardev=chrtpm: tpm-emulator: tpm chardev 'chrtpm' not found. qemu-system-x86_64: -tpmdev emulator,id=tpm0,chardev=chrtpm: tpm-emulator: Could not cleanly shutdown the TPM: No such file or directory QEMU 5.0.90 monitor - type 'help' for more information (qemu) qemu-system-x86_64: -device tpm-tis,tpmdev=tpm0: Property 'tpm-tis.tpmdev' can't find value 'tpm0' $ echo $? 1This is a regression caused by commit
d10e05f15d"tpm: Clean up error reporting intpm_init_tpmdev()". It's incomplete:be->create(opts)continues to useerror_report(), and we don't set an error when it fails.
Quoted 8 lines
Quoted 7 lines
I figure converting the
create()methods to Error would make some sense, but I'm not sure it's worth the effort right now. Revert the broken commit instead, and add a comment totpm_init_tpmdev().Straightforward conflict in
tpm.cresolved.Signed-off-by Markus Armbruster <[email protected]>Reviewed-by Philippe Mathieu-Daudé <[email protected]>
PATCH 2/2 tpm: Improve help on TPM types when none are available
In reply to Markus ArmbrusterHelp is a bit awkward when no TPM types are built into QEMU:
$ upstream-qemu -tpmdev nonexistent,id=tpm0
upstream-qemu: -tpmdev nonexistent,id=tpm0: Parameter 'type' expects a TPM backend type
Supported TPM types (choose only one):Improve to
upstream-qemu: -tpmdev nonexistent,id=tpm0: Parameter 'type' expects a TPM backend type No TPM backend types are available
--- tpm.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-)
tpm.c +9 -4
@@ -47,18 +47,23 @@ tpm_be_find_by_type(enum TpmType type)*/ static void tpm_display_backend_drivers (void ) { +bool got_one = false;int i; -fprintf (stderr,"Supported TPM types (choose only one):\n" ); -for (i =0 ; i <TPM_TYPE__MAX ; i++) {const TPMBackendClass *bc =tpm_be_find_by_type (i);if (!bc) {continue ; } -fprintf (stderr,"%12s %s\n" ,TpmType_str (i), bc->desc); +if (!got_one) { +error_printf ("Supported TPM types (choose only one):\n" ); + got_one = true; + } +error_printf ("%12s %s\n" ,TpmType_str (i), bc->desc); + } +if (!got_one) { +error_printf ("No TPM backend types are available\n" ); } -fprintf (stderr,"\n" ); }/*
Signature · 1 line
2.26.2
Re: [PATCH for-5.1 2/2] tpm: Improve help on TPM types when none are available
Quoted 6 lines · Markus Armbruster
Help is a bit awkward when no TPM types are built into QEMU:
$ upstream-qemu -tpmdev nonexistent,id=tpm0 upstream-qemu: -tpmdev nonexistent,id=tpm0: Parameter 'type' expects a TPM backend type Supported TPM types (choose only one):
Quoted 6 lines
Improve to
upstream-qemu: -tpmdev nonexistent,id=tpm0: Parameter 'type' expects a TPM backend type No TPM backend types are available
Signed-off-by Markus Armbruster <[email protected]>
Quoted 37 lines
--- tpm.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-)tpm.c +9
-4@@ -47,18 +47,23 @@ tpm_be_find_by_type(enum TpmType type)*/ static void tpm_display_backend_drivers (void ) { +bool got_one = false;int i; -fprintf (stderr,"Supported TPM types (choose only one):\n" ); -for (i =0 ; i <TPM_TYPE__MAX ; i++) {const TPMBackendClass *bc =tpm_be_find_by_type (i);if (!bc) {continue ; } -fprintf (stderr,"%12s %s\n" ,TpmType_str (i), bc->desc); +if (!got_one) { +error_printf ("Supported TPM types (choose only one):\n" ); + got_one = true; + } +error_printf ("%12s %s\n" ,TpmType_str (i), bc->desc); + } +if (!got_one) { +error_printf ("No TPM backend types are available\n" ); } -fprintf (stderr,"\n" ); }/*
Stefan Berger
Reviewed-byRe: [PATCH for-5.1 2/2] tpm: Improve help on TPM types when none are available
Quoted 7 lines · Philippe Mathieu-Daudé
Quoted 6 lines · Markus Armbruster
Help is a bit awkward when no TPM types are built into QEMU:
$ upstream-qemu -tpmdev nonexistent,id=tpm0 upstream-qemu: -tpmdev nonexistent,id=tpm0: Parameter 'type' expects a TPM backend type Supported TPM types (choose only one):
Quoted 7 lines
Quoted 6 lines
Improve to
upstream-qemu: -tpmdev nonexistent,id=tpm0: Parameter 'type' expects a TPM backend type No TPM backend types are available
Signed-off-by Markus Armbruster <[email protected]>Reviewed-by Philippe Mathieu-Daudé <[email protected]>
Quoted 38 lines
Quoted 38 lines
--- tpm.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-)tpm.c +9
-4@@ -47,18 +47,23 @@ tpm_be_find_by_type(enum TpmType type)*/ static void tpm_display_backend_drivers (void ) { +bool got_one = false;int i; -fprintf (stderr,"Supported TPM types (choose only one):\n" ); -for (i =0 ; i <TPM_TYPE__MAX ; i++) {const TPMBackendClass *bc =tpm_be_find_by_type (i);if (!bc) {continue ; } -fprintf (stderr,"%12s %s\n" ,TpmType_str (i), bc->desc); +if (!got_one) { +error_printf ("Supported TPM types (choose only one):\n" ); + got_one = true; + } +error_printf ("%12s %s\n" ,TpmType_str (i), bc->desc); + } +if (!got_one) { +error_printf ("No TPM backend types are available\n" ); } -fprintf (stderr,"\n" ); }/*
Re: [PATCH for-5.1 2/2] tpm: Improve help on TPM types when none are available
In reply to Markus ArmbrusterQuoted 3 lines · Markus Armbruster
Help is a bit awkward when no TPM types are built into QEMU:
$ upstream-qemu -tpmdev nonexistent,id=tpm0
I hope you don't mind me replacing 'upstream-qemu' with 'x86_64-softmmu/qemu-system-x86_64'?
Quoted 1 line
upstream-qemu: -tpmdev nonexistent,id=tpm0: Parameter 'type' expects a TPM backend type
and this one with 'qemu-system-x86_64:'
Quoted 6 lines
Supported TPM types (choose only one):
Improve to
upstream-qemu: -tpmdev nonexistent,id=tpm0: Parameter 'type' expects a TPM backend type No TPM backend types are available'
I hope you don't mind me replacing 'upstream-qemu' with 'x86_64-softmmu/qemu-system-x86_64'?
Stefan
Re: [PATCH for-5.1 2/2] tpm: Improve help on TPM types when none are available
Quoted 16 lines · Stefan Berger
Quoted 3 lines · Markus Armbruster
Help is a bit awkward when no TPM types are built into QEMU:
$ upstream-qemu -tpmdev nonexistent,id=tpm0I hope you don't mind me replacing 'upstream-qemu' with 'x86_64-softmmu/qemu-system-x86_64'?
Quoted 1 line
upstream-qemu: -tpmdev nonexistent,id=tpm0: Parameter 'type' expects a TPM backend type
and this one with 'qemu-system-x86_64:'
Quoted 2 lines
Supported TPM types (choose only one):
Quoted 8 lines
Quoted 4 lines
Improve to
upstream-qemu: -tpmdev nonexistent,id=tpm0: Parameter 'type' expects a TPM backend type No TPM backend types are available'
I hope you don't mind me replacing 'upstream-qemu' with 'x86_64-softmmu/qemu-system-x86_64'?
On the contrary! I meant to replace by qemu-system-x86_64 (no directory part), but forgot.
Thanks!