various bugfixes
cover various bugfixes
Collection of bugfixes for 5.0.
Only vhost-user-gpu/grlib_ahb_apb_pnp patches are reviewed.
As 5.0-rc3 is tomorrow, I thought it could help to gather them and resend altogether.
Regards,
Phil.
Mansour Ahmadi (1):
hw/block/pflash: Check return value of blk_pwrite()
Philippe Mathieu-Daudé (11):
Revert "prevent crash when executing guest-file-read with large count"
qga: Extract guest_file_handle_find() to commands-common.h
qga: Extract qmp_guest_file_read() to common commands.c
qga: Restrict guest-file-read count to 48 MB to avoid crashes
vhost-user-gpu: Release memory returned by vu_queue_pop() with free() hw/openrisc/pic_cpu: Use qdev gpio rather than qemu_allocate_irqs() hw/misc/grlib_ahb_apb_pnp: Avoid crash when writing to AHB PnP
registers
hw/misc/grlib_ahb_apb_pnp: Fix AHB PnP 8-bit accesses hw/display/sm501: Avoid heap overflow in sm501_2d_operation()
gdbstub: Do not use memset() on GByteArray
gdbstub: Introduce gdb_get_freg32() to get float32 registers
qga/qapi-schema.json | 6 +++-- include/exec/gdbstub.h | 22 ++++++++++++++++ qga/commands-common.h | 21 +++++++++++++++ contrib/vhost-user-gpu/vhost-user-gpu.c | 4 +-- contrib/vhost-user-gpu/virgl.c | 2 +- hw/block/pflash_cfi01.c | 8 +++++- hw/block/pflash_cfi02.c | 8 +++++- hw/display/sm501.c | 6 +++++ hw/misc/grlib_ahb_apb_pnp.c | 11 ++++++++ hw/openrisc/pic_cpu.c | 5 ++-- qga/commands-posix.c | 29 +++++--------------- qga/commands-win32.c | 35 ++++++------------------- qga/commands.c | 33 +++++++++++++++++++++++ target/arm/gdbstub.c | 3 +-- target/sh4/gdbstub.c | 6 ++--- target/xtensa/gdbstub.c | 6 ++--- 16 files changed, 136 insertions(+), 69 deletions(-) create mode 100644 qga/commands-common.h
Signature · 1 line
2.21.1
PATCH 1/12 Revert "prevent crash when executing guest-file-read with large count"
As noted by Daniel Berrangé in [*], the fix from commit 807e2b6fce which replaced malloc() by try_malloc() is not enough, the process can still run out of memory a few line later:
346 buf = g_try_malloc0(count + 1);
347 if (!buf) {
348 error_setg(errp,
349 "failed to allocate sufficient memory "
350 "to complete the requested service");
351 return NULL;
352 }
353 is_ok = ReadFile(fh, buf, count, &read_count, NULL);
354 if (!is_ok) {
355 error_setg_win32(errp, GetLastError(), "failed to read file");
356 slog("guest-file-read failed, handle %" PRId64, handle);
357 } else {
358 buf[read_count] = 0;
359 read_data = g_new0(GuestFileRead, 1);
^^^^^^Instead we are going to put a low hard limit on 'count' in the next commits. This reverts commit 807e2b6fce022707418bc8f61c069d91c613b3d2.
[*] https://lists.gnu.org/archive/html/qemu-devel/2018-06/msg03471.html
--- qga/commands-win32.c | 8 +------- 1 file changed, 1 insertion(+), 7 deletions(-)
qga/commands-win32.c +1 -7
@@ -343,13 +343,7 @@ GuestFileRead *qmp_guest_file_read(int64_t handle, bool has_count, } fh = gfh->fh; - buf =g_try_malloc0 (count +1 ); -if (!buf) { -error_setg (errp, -"failed to allocate sufficient memory " -"to complete the requested service" ); -return NULL ; - } + buf =g_malloc0 (count +1 ); is_ok =ReadFile (fh, buf, count, &read_count,NULL );if (!is_ok) {error_setg_win32 (errp,GetLastError (),"failed to read file" );
Signature · 1 line
2.21.1
PATCH 2/12 qga: Extract guest_file_handle_find() to commands-common.h
In reply to Philippe Mathieu-DaudéAs we are going to reuse this method, declare it in common header.
--- qga/commands-common.h | 18 ++++++++++++++++++ qga/commands-posix.c | 7 ++++--- qga/commands-win32.c | 7 ++++--- 3 files changed, 26 insertions(+), 6 deletions(-) create mode 100644 qga/commands-common.h
qga/commands-common.h new file +18 -0
@@ -0,0 +1,18 @@ +/* +* QEMU Guest Agent common/cross-platform common commands +* +* Copyright (c) 2020 Red Hat, Inc. +* +* This work is licensed under the terms of the GNU GPL, version 2 or later. +* See the COPYING file in the top-level directory. +*/ +#ifndef QGA_COMMANDS_COMMON_H +#define QGA_COMMANDS_COMMON_H + +#include "qga-qapi-types.h" + +typedef struct GuestFileHandle GuestFileHandle ; + +GuestFileHandle *guest_file_handle_find (int64_t id,Error **errp); + +#endif
qga/commands-posix.c +4 -3
@@ -26,6 +26,7 @@#include "qemu/sockets.h" #include "qemu/base64.h" #include "qemu/cutils.h" +#include "commands-common.h" #ifdef HAVE_UTMPX #include <utmpx.h>
@@ -237,12 +238,12 @@ typedef enum {RW_STATE_WRITING , } RwState; -typedef struct GuestFileHandle { +struct GuestFileHandle {uint64_t id;FILE *fh;RwState state;QTAILQ_ENTRY (GuestFileHandle) next; -}GuestFileHandle ; +}; static struct { QTAILQ_HEAD(, GuestFileHandle) filehandles;
@@ -268,7 +269,7 @@ static int64_t guest_file_handle_add(FILE *fh, Error **errp)return handle; } -static GuestFileHandle *guest_file_handle_find (int64_t id,Error **errp) +GuestFileHandle *guest_file_handle_find (int64_t id,Error **errp) {GuestFileHandle *gfh;
qga/commands-win32.c +4 -3
@@ -37,6 +37,7 @@#include "qemu/queue.h" #include "qemu/host-utils.h" #include "qemu/base64.h" +#include "commands-common.h" #ifndef SHTDN_REASON_FLAG_PLANNED #define SHTDN_REASON_FLAG_PLANNED 0x80000000
@@ -50,11 +51,11 @@#define INVALID_SET_FILE_POINTER ((DWORD)-1) -typedef struct GuestFileHandle { +struct GuestFileHandle {int64_t id;HANDLE fh;QTAILQ_ENTRY (GuestFileHandle) next; -}GuestFileHandle ; +}; static struct { QTAILQ_HEAD(, GuestFileHandle) filehandles;
@@ -126,7 +127,7 @@ static int64_t guest_file_handle_add(HANDLE fh, Error **errp)return handle; } -static GuestFileHandle *guest_file_handle_find (int64_t id,Error **errp) +GuestFileHandle *guest_file_handle_find (int64_t id,Error **errp) {GuestFileHandle *gfh; QTAILQ_FOREACH(gfh, &guest_file_state.filehandles, next) {
Signature · 1 line
2.21.1
PATCH 3/12 qga: Extract qmp_guest_file_read() to common commands.c
In reply to Philippe Mathieu-DaudéExtract the common code shared by both POSIX/Win32 implementations.
--- qga/commands-common.h | 3 +++ qga/commands-posix.c | 22 +++------------------- qga/commands-win32.c | 20 +++----------------- qga/commands.c | 26 ++++++++++++++++++++++++++ 4 files changed, 35 insertions(+), 36 deletions(-)
qga/commands-common.h +3 -0
@@ -15,4 +15,7 @@ typedef struct GuestFileHandle GuestFileHandle;GuestFileHandle *guest_file_handle_find (int64_t id,Error **errp); +GuestFileRead *guest_file_read_unsafe (GuestFileHandle *gfh, +int64_t count,Error **errp); +#endif
qga/commands-posix.c +3 -19
@@ -461,29 +461,14 @@ void qmp_guest_file_close(int64_t handle, Error **errp)g_free (gfh); } -struct GuestFileRead *qmp_guest_file_read (int64_t handle,bool has_count, -int64_t count,Error **errp) +GuestFileRead *guest_file_read_unsafe (GuestFileHandle *gfh, +int64_t count,Error **errp) { -GuestFileHandle *gfh =guest_file_handle_find (handle, errp);GuestFileRead *read_data =NULL ;guchar *buf; -FILE *fh; +FILE *fh = gfh->fh;size_t read_count; -if (!gfh) { -return NULL ; - } - -if (!has_count) { - count =QGA_READ_COUNT_DEFAULT ; - }else if (count <0 || count >=UINT32_MAX ) { -error_setg (errp,"value '%" PRId64"' is invalid for argument count" , - count); -return NULL ; - } - - fh = gfh->fh; -/* explicitly flush when switching from writing to reading */ if (gfh->state ==RW_STATE_WRITING ) {int ret =fflush (fh);
@@ -498,7 +483,6 @@ struct GuestFileRead *qmp_guest_file_read(int64_t handle, bool has_count, read_count =fread (buf,1 , count, fh);if (ferror (fh)) {error_setg_errno (errp, errno,"failed to read file" ); -slog ("guest-file-read failed, handle: %" PRId64, handle); }else { buf[read_count] =0 ; read_data =g_new0 (GuestFileRead,1 );
qga/commands-win32.c +3 -17
@@ -322,33 +322,19 @@ void qmp_guest_shutdown(bool has_mode, const char *mode, Error **errp) } } -GuestFileRead *qmp_guest_file_read (int64_t handle,bool has_count, -int64_t count,Error **errp) +GuestFileRead *guest_file_read_unsafe (GuestFileHandle *gfh, +int64_t count,Error **errp) {GuestFileRead *read_data =NULL ;guchar *buf; -HANDLE fh; +HANDLE fh = gfh->fh;bool is_ok;DWORD read_count; -GuestFileHandle *gfh =guest_file_handle_find (handle, errp); -if (!gfh) { -return NULL ; - } -if (!has_count) { - count =QGA_READ_COUNT_DEFAULT ; - }else if (count <0 || count >=UINT32_MAX ) { -error_setg (errp,"value '%" PRId64 -"' is invalid for argument count" , count); -return NULL ; - } - - fh = gfh->fh; buf =g_malloc0 (count +1 ); is_ok =ReadFile (fh, buf, count, &read_count,NULL );if (!is_ok) {error_setg_win32 (errp,GetLastError (),"failed to read file" ); -slog ("guest-file-read failed, handle %" PRId64, handle); }else { buf[read_count] =0 ; read_data =g_new0 (GuestFileRead,1 );
qga/commands.c +26 -0
@@ -18,6 +18,7 @@#include "qemu/base64.h" #include "qemu/cutils.h" #include "qemu/atomic.h" +#include "commands-common.h" /* Maximum captured guest-exec out_data/err_data - 16MB */ #define GUEST_EXEC_MAX_OUTPUT (16*1024*1024)
@@ -547,3 +548,28 @@ error:g_free (info);return NULL ; } + +GuestFileRead *qmp_guest_file_read (int64_t handle,bool has_count, +int64_t count,Error **errp) +{ +GuestFileHandle *gfh =guest_file_handle_find (handle, errp); +GuestFileRead *read_data; + +if (!gfh) { +return NULL ; + } +if (!has_count) { + count =QGA_READ_COUNT_DEFAULT ; + }else if (count <0 || count >=UINT32_MAX ) { +error_setg (errp,"value '%" PRId64"' is invalid for argument count" , + count); +return NULL ; + } + + read_data =guest_file_read_unsafe (gfh, count, errp); +if (!read_data) { +slog ("guest-file-write failed, handle: %" PRId64, handle); + } + +return read_data; +}
Signature · 1 line
2.21.1
PATCH 4/12 qga: Restrict guest-file-read count to 48 MB to avoid crashes
In reply to Philippe Mathieu-DaudéOn [*] Daniel Berrangé commented:
The QEMU guest agent protocol is not sensible way to access huge files inside the guest. It requires the inefficient process of reading the entire data into memory than duplicating it again in base64 format, and then copying it again in the JSON serializer / monitor code.
For arbitrary general purpose file access, especially for large files, use a real file transfer program or use a network block device, not the QEMU guest agent.
To avoid bug reports as BZ#1594054 (CVE-2018-12617), follow his suggestion to put a low, hard limit on "count" in the guest agent QAPI schema, and don't allow count to be larger than 48 MB.
[*] https://www.mail-archive.com/[email protected]/msg693176.html
--- qga/qapi-schema.json | 6 ++++-- qga/commands.c | 9 ++++++++- 2 files changed, 12 insertions(+), 3 deletions(-)
qga/qapi-schema.json +4 -2
@@ -266,11 +266,13 @@ ## # @guest-file-read: # -# Read from an open file in the guest. Data will be base64-encoded +# Read from an open file in the guest. Data will be base64-encoded. +# As this command is just for limited, ad-hoc debugging, such as log +# file access, the number of bytes to read is limited to 10 MB. # # @handle: filehandle returned by guest-file-open # -# @count: maximum number of bytes to read (default is 4KB) +# @count: maximum number of bytes to read (default is 4KB, maximum is 10MB) # # Returns: @GuestFileRead on success. #
qga/commands.c +8 -1
@@ -11,6 +11,7 @@*/ #include "qemu/osdep.h" +#include "qemu/units.h" #include "guest-agent-core.h" #include "qga-qapi-commands.h" #include "qapi/error.h"
@@ -24,6 +25,12 @@#define GUEST_EXEC_MAX_OUTPUT (16*1024*1024)/* Allocation and I/O buffer for reading guest-exec out_data/err_data - 4KB */ #define GUEST_EXEC_IO_SIZE (4*1024) +/* +* Maximum file size to read - 48MB +* +* (48MB + Base64 3:4 overhead = JSON parser 64 MB limit) +*/ +#define GUEST_FILE_READ_COUNT_MAX (48 * MiB)/* Note: in some situations, like with the fsfreeze, logging may be * temporarilly disabled. if it is necessary that a command be able
@@ -560,7 +567,7 @@ GuestFileRead *qmp_guest_file_read(int64_t handle, bool has_count, }if (!has_count) { count =QGA_READ_COUNT_DEFAULT ; - }else if (count <0 || count >=UINT32_MAX ) { + }else if (count <0 || count >GUEST_FILE_READ_COUNT_MAX ) {error_setg (errp,"value '%" PRId64"' is invalid for argument count" , count);return NULL ;
Signature · 1 line
2.21.1
Re: [PATCH-for-5.0 04/12] qga: Restrict guest-file-read count to 48 MB to avoid crashes
Quoted 33 lines · Philippe Mathieu-Daudé
On [*] Daniel Berrangé commented:
The QEMU guest agent protocol is not sensible way to access huge files inside the guest. It requires the inefficient process of reading the entire data into memory than duplicating it again in base64 format, and then copying it again in the JSON serializer / monitor code.
For arbitrary general purpose file access, especially for large files, use a real file transfer program or use a network block device, not the QEMU guest agent.
To avoid bug reports as BZ#1594054 (CVE-2018-12617), follow his suggestion to put a low, hard limit on "count" in the guest agent QAPI schema, and don't allow count to be larger than 48 MB.
[*] https://www.mail-archive.com/[email protected]/msg693176.html
Fixes CVE-2018-12617Reported-by Fakhri Zulkifli <[email protected]>Suggested-by Daniel P. Berrangé <[email protected]>Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- qga/qapi-schema.json | 6 ++++-- qga/commands.c | 9 ++++++++- 2 files changed, 12 insertions(+), 3 deletions(-)qga/qapi-schema.json +0
-0
qga/qapi-schema.json line 271
Quoted 7 lines
qga/qapi-schema.json +3
-1@@ -266,11 +266,13 @@ ## # @guest-file-read: # -# Read from an open file in the guest. Data will be base64-encoded +# Read from an open file in the guest. Data will be base64-encoded. +# As this command is just for limited, ad-hoc debugging, such as log +# file access, the number of bytes to read is limited to 10 MB.
s/10/48/
qga/qapi-schema.json line 275
Quoted 5 lines
qga/qapi-schema.json +1
-1# # @handle: filehandle returned by guest-file-open # -# @count: maximum number of bytes to read (default is 4KB) +# @count: maximum number of bytes to read (default is 4KB, maximum is 10MB)
s/10/48/
Quoted 29 lines
qga/qapi-schema.json +0
-0# # Returns: @GuestFileRead on success. #qga/commands.c +7
-0@@ -11,6 +11,7 @@*/ #include "qemu/osdep.h" +#include "qemu/units.h" #include "guest-agent-core.h" #include "qga-qapi-commands.h" #include "qapi/error.h" @@ -24,6 +25,12 @@#define GUEST_EXEC_MAX_OUTPUT (16*1024*1024)/* Allocation and I/O buffer for reading guest-exec out_data/err_data - 4KB */ #define GUEST_EXEC_IO_SIZE (4*1024) +/* +* Maximum file size to read - 48MB +* +* (48MB + Base64 3:4 overhead = JSON parser 64 MB limit) +*/ +#define GUEST_FILE_READ_COUNT_MAX (48 * MiB)/* Note: in some situations, like with the fsfreeze, logging may be * temporarilly disabled. if it is necessary that a command be able
qga/commands.c GuestFileRead *qmp_guest_file_read(int64_t handle, bool has_count, · line 573
Quoted 8 lines
qga/commands.c +1
-1@@ -560,7 +567,7 @@ GuestFileRead *qmp_guest_file_read(int64_t handle, bool has_count, }if (!has_count) { count =QGA_READ_COUNT_DEFAULT ; - }else if (count <0 || count >=UINT32_MAX ) { + }else if (count <0 || count >GUEST_FILE_READ_COUNT_MAX ) {error_setg (errp,"value '%" PRId64"' is invalid for argument count" , count);return NULL ;
With the docs typos fixed:
Regards,
Daniel
Signature · 3 lines
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
Re: [PATCH-for-5.0 04/12] qga: Restrict guest-file-read count to 48 MB to avoid crashes
Quoted 44 lines · Daniel P. Berrangé
Quoted 33 lines · Philippe Mathieu-Daudé
On [*] Daniel Berrangé commented:
The QEMU guest agent protocol is not sensible way to access huge files inside the guest. It requires the inefficient process of reading the entire data into memory than duplicating it again in base64 format, and then copying it again in the JSON serializer / monitor code.
For arbitrary general purpose file access, especially for large files, use a real file transfer program or use a network block device, not the QEMU guest agent.
To avoid bug reports as BZ#1594054 (CVE-2018-12617), follow his suggestion to put a low, hard limit on "count" in the guest agent QAPI schema, and don't allow count to be larger than 48 MB.
[*] https://www.mail-archive.com/[email protected]/msg693176.html
Fixes CVE-2018-12617Reported-by Fakhri Zulkifli <[email protected]>Suggested-by Daniel P. Berrangé <[email protected]>Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- qga/qapi-schema.json | 6 ++++-- qga/commands.c | 9 ++++++++- 2 files changed, 12 insertions(+), 3 deletions(-)qga/qapi-schema.json +0
-0qga/qapi-schema.json line 271
Quoted 7 lines
qga/qapi-schema.json +3
-1@@ -266,11 +266,13 @@ ## # @guest-file-read: # -# Read from an open file in the guest. Data will be base64-encoded +# Read from an open file in the guest. Data will be base64-encoded. +# As this command is just for limited, ad-hoc debugging, such as log +# file access, the number of bytes to read is limited to 10 MB.s/10/48/
Quoted 7 lines
qga/qapi-schema.json line 275
Quoted 5 lines
qga/qapi-schema.json +1
-1# # @handle: filehandle returned by guest-file-open # -# @count: maximum number of bytes to read (default is 4KB) +# @count: maximum number of bytes to read (default is 4KB, maximum is 10MB)s/10/48/
Oops I totally missed those, thanks!
Quoted 47 lines
Quoted 30 lines
qga/qapi-schema.json +0
-0# # Returns: @GuestFileRead on success. #qga/commands.c +7
-0@@ -11,6 +11,7 @@*/ #include "qemu/osdep.h" +#include "qemu/units.h" #include "guest-agent-core.h" #include "qga-qapi-commands.h" #include "qapi/error.h" @@ -24,6 +25,12 @@#define GUEST_EXEC_MAX_OUTPUT (16*1024*1024)/* Allocation and I/O buffer for reading guest-exec out_data/err_data - 4KB */ #define GUEST_EXEC_IO_SIZE (4*1024) +/* +* Maximum file size to read - 48MB +* +* (48MB + Base64 3:4 overhead = JSON parser 64 MB limit) +*/ +#define GUEST_FILE_READ_COUNT_MAX (48 * MiB)/* Note: in some situations, like with the fsfreeze, logging may be * temporarilly disabled. if it is necessary that a command be able qga/commands.c GuestFileRead *qmp_guest_file_read(int64_t handle, bool has_count, · line 573
Quoted 8 lines
qga/commands.c +1
-1@@ -560,7 +567,7 @@ GuestFileRead *qmp_guest_file_read(int64_t handle, bool has_count, }if (!has_count) { count =QGA_READ_COUNT_DEFAULT ; - }else if (count <0 || count >=UINT32_MAX ) { + }else if (count <0 || count >GUEST_FILE_READ_COUNT_MAX ) {error_setg (errp,"value '%" PRId64"' is invalid for argument count" , count);return NULL ;With the docs typos fixed:
Reviewed-by Daniel P. Berrangé <[email protected]>Regards,
Daniel
Re: [PATCH-for-5.0 04/12] qga: Restrict guest-file-read count to 48 MB to avoid crashes
Quoting Philippe Mathieu-Daudé (2020-04-15 08:02:18)
Quoted 53 lines · Daniel P. Berrangé
Quoted 44 lines · Philippe Mathieu-Daudé
Quoted 33 lines
On [*] Daniel Berrangé commented:
The QEMU guest agent protocol is not sensible way to access huge files inside the guest. It requires the inefficient process of reading the entire data into memory than duplicating it again in base64 format, and then copying it again in the JSON serializer / monitor code.
For arbitrary general purpose file access, especially for large files, use a real file transfer program or use a network block device, not the QEMU guest agent.
To avoid bug reports as BZ#1594054 (CVE-2018-12617), follow his suggestion to put a low, hard limit on "count" in the guest agent QAPI schema, and don't allow count to be larger than 48 MB.
[*] https://www.mail-archive.com/[email protected]/msg693176.html
Fixes CVE-2018-12617Reported-by Fakhri Zulkifli <[email protected]>Suggested-by Daniel P. Berrangé <[email protected]>Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- qga/qapi-schema.json | 6 ++++-- qga/commands.c | 9 ++++++++- 2 files changed, 12 insertions(+), 3 deletions(-)qga/qapi-schema.json +0
-0qga/qapi-schema.json line 271
Quoted 7 lines
qga/qapi-schema.json +3
-1@@ -266,11 +266,13 @@ ## # @guest-file-read: # -# Read from an open file in the guest. Data will be base64-encoded +# Read from an open file in the guest. Data will be base64-encoded. +# As this command is just for limited, ad-hoc debugging, such as log +# file access, the number of bytes to read is limited to 10 MB.s/10/48/
Quoted 7 lines
qga/qapi-schema.json line 275
Quoted 5 lines
qga/qapi-schema.json +1
-1# # @handle: filehandle returned by guest-file-open # -# @count: maximum number of bytes to read (default is 4KB) +# @count: maximum number of bytes to read (default is 4KB, maximum is 10MB)s/10/48/
Quoted 1 line
Oops I totally missed those, thanks!
I've rolled in the doc fix-ups and sent a pull request for patches 1-4
Quoted 49 lines
Quoted 48 lines
Quoted 31 lines
qga/qapi-schema.json +0
-0# # Returns: @GuestFileRead on success. #qga/commands.c +7
-0@@ -11,6 +11,7 @@*/ #include "qemu/osdep.h" +#include "qemu/units.h" #include "guest-agent-core.h" #include "qga-qapi-commands.h" #include "qapi/error.h" @@ -24,6 +25,12 @@#define GUEST_EXEC_MAX_OUTPUT (16*1024*1024)/* Allocation and I/O buffer for reading guest-exec out_data/err_data - 4KB */ #define GUEST_EXEC_IO_SIZE (4*1024) +/* +* Maximum file size to read - 48MB +* +* (48MB + Base64 3:4 overhead = JSON parser 64 MB limit) +*/ +#define GUEST_FILE_READ_COUNT_MAX (48 * MiB)/* Note: in some situations, like with the fsfreeze, logging may be * temporarilly disabled. if it is necessary that a command be able qga/commands.c GuestFileRead *qmp_guest_file_read(int64_t handle, bool has_count, · line 573
Quoted 8 lines
qga/commands.c +1
-1@@ -560,7 +567,7 @@ GuestFileRead *qmp_guest_file_read(int64_t handle, bool has_count, }if (!has_count) { count =QGA_READ_COUNT_DEFAULT ; - }else if (count <0 || count >=UINT32_MAX ) { + }else if (count <0 || count >GUEST_FILE_READ_COUNT_MAX ) {error_setg (errp,"value '%" PRId64"' is invalid for argument count" , count);return NULL ;With the docs typos fixed:
Reviewed-by Daniel P. Berrangé <[email protected]>Regards,
Daniel
PATCH 5/12 vhost-user-gpu: Release memory returned by vu_queue_pop() with free()
In reply to Philippe Mathieu-Daudévu_queue_pop() returns memory that must be freed with free().
--- contrib/vhost-user-gpu/vhost-user-gpu.c | 4 ++-- contrib/vhost-user-gpu/virgl.c | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-)
contrib/vhost-user-gpu/vhost-user-gpu.c +2 -2
@@ -848,7 +848,7 @@ vg_handle_ctrl(VuDev *dev, int qidx)QTAILQ_INSERT_TAIL (&vg->fenceq, cmd, next); vg->inflight++; }else { -g_free (cmd); +free (cmd); } } }
@@ -939,7 +939,7 @@ vg_handle_cursor(VuDev *dev, int qidx) }vu_queue_push (dev, vq, elem,0 );vu_queue_notify (dev, vq); -g_free (elem); +free (elem); } }
contrib/vhost-user-gpu/virgl.c +1 -1
@@ -519,7 +519,7 @@ virgl_write_fence(void *opaque, uint32_t fence)g_debug ("FENCE %" PRIu64, cmd->cmd_hdr.fence_id);vg_ctrl_response_nodata (g, cmd,VIRTIO_GPU_RESP_OK_NODATA );QTAILQ_REMOVE (&g->fenceq, cmd, next); -g_free (cmd); +free (cmd); g->inflight--; } }
Signature · 1 line
2.21.1
Re: [PATCH-for-5.0 05/12] vhost-user-gpu: Release memory returned by vu_queue_pop() with free()
Quoted 7 lines · Philippe Mathieu-Daudé
vu_queue_pop()returns memory that must be freed withfree().Reported-by Coverity (CID 1421887 ALLOC_FREE_MISMATCH)Suggested-by Peter Maydell <[email protected]>Reviewed-by Marc-André Lureau <[email protected]>Signed-off-by Philippe Mathieu-Daudé <[email protected]>
Quoted 42 lines
--- contrib/vhost-user-gpu/vhost-user-gpu.c | 4 ++-- contrib/vhost-user-gpu/virgl.c | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-)contrib/vhost-user-gpu/vhost-user-gpu.c +2
-2@@ -848,7 +848,7 @@ vg_handle_ctrl(VuDev *dev, int qidx)QTAILQ_INSERT_TAIL (&vg->fenceq, cmd, next); vg->inflight++; }else { -g_free (cmd); +free (cmd); } } }@@ -939,7 +939,7 @@ vg_handle_cursor(VuDev *dev, int qidx) }vu_queue_push (dev, vq, elem,0 );vu_queue_notify (dev, vq); -g_free (elem); +free (elem); } }contrib/vhost-user-gpu/virgl.c +1
-1@@ -519,7 +519,7 @@ virgl_write_fence(void *opaque, uint32_t fence)g_debug ("FENCE %" PRIu64, cmd->cmd_hdr.fence_id);vg_ctrl_response_nodata (g, cmd,VIRTIO_GPU_RESP_OK_NODATA );QTAILQ_REMOVE (&g->fenceq, cmd, next); -g_free (cmd); +free (cmd); g->inflight--; } }Signature · 1 line
2.21.1
PATCH 6/12 hw/openrisc/pic_cpu: Use qdev gpio rather than qemu_allocate_irqs()
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
Coverity points out (CID 1421934) that we are leaking the memory returned by qemu_allocate_irqs(). We can avoid this leak by switching to using qdev_init_gpio_in(); the base class finalize will free the irqs that this allocates under the hood.
Patch created mechanically using spatch with this script inspired from commit d6ef883d9d7:
@@
typedef qemu_irq;
identifier irqs, handler;
expression opaque, count, i;
@@
- qemu_irq *irqs;
...
- irqs = qemu_allocate_irqs(handler, opaque, count);
+ qdev_init_gpio_in(DEVICE(opaque), handler, count);
<+...
- irqs[i]
+ qdev_get_gpio_in(DEVICE(opaque), i)
...+> ?- g_free(irqs);
Message-Id: <[email protected]>
--- hw/openrisc/pic_cpu.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)
hw/openrisc/pic_cpu.c +2 -3
@@ -52,10 +52,9 @@ static void openrisc_pic_cpu_handler(void *opaque, int irq, int level)void cpu_openrisc_pic_init (OpenRISCCPU *cpu) {int i; -qemu_irq *qi; - qi =qemu_allocate_irqs (openrisc_pic_cpu_handler, cpu,NR_IRQS ); +qdev_init_gpio_in (DEVICE (cpu), openrisc_pic_cpu_handler,NR_IRQS );for (i =0 ; i <NR_IRQS ; i++) { - cpu->env.irq[i] = qi[i]; + cpu->env.irq[i] =qdev_get_gpio_in (DEVICE (cpu), i); } }
Signature · 1 line
2.21.1
PATCH 7/12 hw/misc/grlib_ahb_apb_pnp: Avoid crash when writing to AHB PnP registers
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
Similarly to commit 158b659451 with the APB PnP registers, guests can crash QEMU when writting to the AHB PnP registers:
$ echo 'writeb 0xfffff042 69' | qemu-system-sparc -M leon3_generic -S -bios /etc/magic -qtest stdio [I 1571938309.932255] OPENED [R +0.063474] writeb 0xfffff042 69 Segmentation fault (core dumped)
(gdb) bt
#0 0x0000000000000000 in ()
#1 0x0000562999110df4 in memory_region_write_with_attrs_accessor
(mr=mr@entry=0x56299aa28ea0, addr=66, value=value@entry=0x7fff6abe13b8, size=size@entry=1, shift=<optimized out>, mask=mask@entry=255, attrs=...) at memory.c:503
#2 0x000056299911095e in access_with_adjusted_size
(addr=addr@entry=66, value=value@entry=0x7fff6abe13b8, size=size@entry=1, access_size_min=<optimized out>, access_size_max=<optimized out>, access_fn=access_fn@entry=
0x562999110d70 <memory_region_write_with_attrs_accessor>, mr=0x56299aa28ea0, attrs=...) at memory.c:539
#3 0x0000562999114fba in memory_region_dispatch_write (mr=mr@entry=0x56299aa28ea0, addr=66, data=<optimized out>, op=<optimized out>, attrs=attrs@entry=...) at memory.c:1482
#4 0x00005629990c0860 in flatview_write_continue
(fv=fv@entry=0x56299aa7d8a0, addr=addr@entry=4294963266, attrs=..., ptr=ptr@entry=0x7fff6abe1540, len=len@entry=1, addr1=<optimized out>, l=<optimized out>, mr=0x56299aa28ea0)
at include/qemu/host-utils.h:164
#5 0x00005629990c0a76 in flatview_write (fv=0x56299aa7d8a0, addr=4294963266, attrs=..., buf=0x7fff6abe1540, len=1) at exec.c:3165
#6 0x00005629990c4c1b in address_space_write (as=<optimized out>, addr=<optimized out>, attrs=..., attrs@entry=..., buf=buf@entry=0x7fff6abe1540, len=len@entry=1) at exec.c:3256
#7 0x000056299910f807 in qtest_process_command (chr=chr@entry=0x5629995ee920 <qtest_chr>, words=words@entry=0x56299acfcfa0) at qtest.c:437Instead of crashing, log the access as unimplemented.
Message-Id: <[email protected]>
--- hw/misc/grlib_ahb_apb_pnp.c | 7 +++++++ 1 file changed, 7 insertions(+)
hw/misc/grlib_ahb_apb_pnp.c +7 -0
@@ -136,8 +136,15 @@ static uint64_t grlib_ahb_pnp_read(void *opaque, hwaddr offset, unsigned size)return ahb_pnp->regs[offset >>2 ]; } +static void grlib_ahb_pnp_write (void *opaque,hwaddr addr, +uint64_t val,unsigned size) +{ +qemu_log_mask (LOG_UNIMP ,"%s not implemented\n" , __func__); +} +static const MemoryRegionOps grlib_ahb_pnp_ops = { .read = grlib_ahb_pnp_read, + .write = grlib_ahb_pnp_write, .endianness =DEVICE_BIG_ENDIAN , };
Signature · 1 line
2.21.1
PATCH 8/12 hw/misc/grlib_ahb_apb_pnp: Fix AHB PnP 8-bit accesses
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
The Plug & Play region of the AHB/APB bridge can be accessed by various word size, however the implementation is clearly restricted to 32-bit:
static uint64_t grlib_ahb_pnp_read(void *opaque, hwaddr offset, unsigned size)
{
AHBPnp *ahb_pnp = GRLIB_AHB_PNP(opaque);return ahb_pnp->regs[offset >> 2];
}
Similarly to commit 0fbe394a64 with the APB PnP registers, set the MemoryRegionOps::impl min/max fields to 32-bit, so memory.c::access_with_adjusted_size() can adjust when the access is not 32-bit.
Message-Id: <[email protected]>
--- hw/misc/grlib_ahb_apb_pnp.c | 4 ++++ 1 file changed, 4 insertions(+)
hw/misc/grlib_ahb_apb_pnp.c +4 -0
@@ -146,6 +146,10 @@ static const MemoryRegionOps grlib_ahb_pnp_ops = { .read = grlib_ahb_pnp_read, .write = grlib_ahb_pnp_write, .endianness = DEVICE_BIG_ENDIAN, + .impl = { + .min_access_size = 4, + .max_access_size = 4, + }, };static void grlib_ahb_pnp_realize (DeviceState *dev,Error **errp)
Signature · 1 line
2.21.1
PATCH 9/12 hw/display/sm501: Avoid heap overflow in sm501_2d_operation()
In reply to Philippe Mathieu-DaudéFrom: Philippe Mathieu-Daudé <[email protected]>
Zhang Zi Ming reported a heap overflow in the Drawing Engine of the SM501 companion chip model, in particular in the COPY_AREA() macro in sm501_2d_operation().
Add a simple check to avoid the heap overflow.
This fixes:
=================================================================
==20518==ERROR: AddressSanitizer: heap-buffer-overflow on address 0x7f6f4c3fffff at pc 0x55b1e1d358f0 bp 0x7ffce464dfb0 sp 0x7ffce464dfa8 READ of size 1 at 0x7f6f4c3fffff thread T0
#0 0x55b1e1d358ef in sm501_2d_operation hw/display/sm501.c:788:13
#1 0x55b1e1d32c38 in sm501_2d_engine_write hw/display/sm501.c:1466:13
#2 0x55b1e0cd19d8 in memory_region_write_accessor memory.c:483:5
#3 0x55b1e0cd1404 in access_with_adjusted_size memory.c:544:18
#4 0x55b1e0ccfb9d in memory_region_dispatch_write memory.c:1476:16
#5 0x55b1e0ae55a8 in flatview_write_continue exec.c:3125:23
#6 0x55b1e0ad3e87 in flatview_write exec.c:3165:14
#7 0x55b1e0ad3a24 in address_space_write exec.c:3256:18 0x7f6f4c3fffff is located 4194303 bytes to the right of 4194304-byte region [0x7f6f4bc00000,0x7f6f4c000000) allocated by thread T0 here:
#0 0x55b1e0a6e715 in __interceptor_posix_memalign (ppc64-softmmu/qemu-system-ppc64+0x19c0715)
#1 0x55b1e31c1482 in qemu_try_memalign util/oslib-posix.c:189:11
#2 0x55b1e31c168c in qemu_memalign util/oslib-posix.c:205:27
#3 0x55b1e11a00b3 in spapr_reallocate_hpt hw/ppc/spapr.c:1560:23
#4 0x55b1e11a0ce4 in spapr_setup_hpt hw/ppc/spapr.c:1593:5
#5 0x55b1e11c2fba in spapr_machine_reset hw/ppc/spapr.c:1644:9
#6 0x55b1e1368b01 in qemu_system_reset softmmu/vl.c:1391:9
#7 0x55b1e1375af3 in qemu_init softmmu/vl.c:4436:5
#8 0x55b1e2fc8a59 in main softmmu/main.c:48:5
#9 0x7f6f8150bf42 in __libc_start_main (/lib64/libc.so.6+0x23f42) SUMMARY: AddressSanitizer: heap-buffer-overflow hw/display/sm501.c:788:13 in sm501_2d_operation Shadow bytes around the buggy address:
0x0fee69877fa0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x0fee69877fb0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x0fee69877fc0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x0fee69877fd0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x0fee69877fe0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
=>0x0fee69877ff0: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa[fa]
0x0fee69878000: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
0x0fee69878010: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
0x0fee69878020: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
0x0fee69878030: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
0x0fee69878040: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 Shadow byte legend (one shadow byte represents 8 application bytes):
Addressable: 00
Partially addressable: 01 02 03 04 05 06 07
Heap left redzone: fa
Freed heap region: fd
Poisoned by user: f7
ASan internal: fe
==20518==ABORTING
Message-Id: <[email protected]>
--- hw/display/sm501.c | 6 ++++++ 1 file changed, 6 insertions(+)
hw/display/sm501.c +6 -0
@@ -726,6 +726,12 @@ static void sm501_2d_operation(SM501State *s)int crt = (s->dc_crt_control &SM501_DC_CRT_CONTROL_SEL ) ?1 :0 ;int fb_len =get_width (s, crt) *get_height (s, crt) *get_bpp (s, crt); +if (rtl && (src_x < operation_width || src_y < operation_height)) { +qemu_log_mask (LOG_GUEST_ERROR ,"sm501: Illegal RTL address (%i, %i)\n" , + src_x, src_y); +return ; + } +if (addressing !=0x0 ) {printf ("%s: only XY addressing is supported.\n" , __func__);abort ();
Signature · 1 line
2.21.1
PATCH 10/12 hw/block/pflash: Check return value of blk_pwrite()
In reply to Philippe Mathieu-DaudéFrom: Mansour Ahmadi <[email protected]>
When updating the PFLASH file contents, we should check for a possible failure of blk_pwrite(). Similar to commit 3a688294e.
Message-Id: <[email protected]>
--- hw/block/pflash_cfi01.c | 8 +++++++- hw/block/pflash_cfi02.c | 8 +++++++- 2 files changed, 14 insertions(+), 2 deletions(-)
hw/block/pflash_cfi01.c +7 -1
@@ -42,6 +42,7 @@#include "hw/qdev-properties.h" #include "sysemu/block-backend.h" #include "qapi/error.h" +#include "qemu/error-report.h" #include "qemu/timer.h" #include "qemu/bitops.h" #include "qemu/error-report.h"
@@ -399,13 +400,18 @@ static void pflash_update(PFlashCFI01 *pfl, int offset, int size) {int offset_end; +int ret;if (pfl->blk) { offset_end = offset + size;/* widen to sector boundaries */ offset =QEMU_ALIGN_DOWN (offset,BDRV_SECTOR_SIZE ); offset_end =QEMU_ALIGN_UP (offset_end,BDRV_SECTOR_SIZE ); -blk_pwrite (pfl->blk, offset, pfl->storage + offset, + ret =blk_pwrite (pfl->blk, offset, pfl->storage + offset, offset_end - offset,0 ); +if (ret <0 ) { +/* TODO set error bit in status */ +error_report ("Could not update PFLASH: %s" ,strerror (-ret)); + } } }
hw/block/pflash_cfi02.c +7 -1
@@ -37,6 +37,7 @@#include "hw/block/flash.h" #include "hw/qdev-properties.h" #include "qapi/error.h" +#include "qemu/error-report.h" #include "qemu/bitmap.h" #include "qemu/timer.h" #include "sysemu/block-backend.h"
@@ -393,13 +394,18 @@ static uint64_t pflash_read(void *opaque, hwaddr offset, unsigned int width)static void pflash_update (PFlashCFI02 *pfl,int offset,int size) {int offset_end; +int ret;if (pfl->blk) { offset_end = offset + size;/* widen to sector boundaries */ offset =QEMU_ALIGN_DOWN (offset,BDRV_SECTOR_SIZE ); offset_end =QEMU_ALIGN_UP (offset_end,BDRV_SECTOR_SIZE ); -blk_pwrite (pfl->blk, offset, pfl->storage + offset, + ret =blk_pwrite (pfl->blk, offset, pfl->storage + offset, offset_end - offset,0 ); +if (ret <0 ) { +/* TODO set error bit in status */ +error_report ("Could not update PFLASH: %s" ,strerror (-ret)); + } } }
Signature · 1 line
2.21.1
Re: [PATCH-for-5.0 10/12] hw/block/pflash: Check return value of blk_pwrite()
Thank you for fixing the patch, Philippe!
Quoted 84 lines · Philippe Mathieu-Daudé
From: Mansour Ahmadi <[email protected]>
When updating the PFLASH file contents, we should check for a possible failure of
blk_pwrite(). Similar to commit3a688294e.Signed-off-by Mansour Ahmadi <[email protected]>Message-Id: <[email protected]>
[PMD: Add missing "qemu/error-report.h" include and TODO comment]Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- hw/block/pflash_cfi01.c | 8 +++++++- hw/block/pflash_cfi02.c | 8 +++++++- 2 files changed, 14 insertions(+), 2 deletions(-)hw/block/pflash_cfi01.c +7
-1@@ -42,6 +42,7 @@#include "hw/qdev-properties.h" #include "sysemu/block-backend.h" #include "qapi/error.h" +#include "qemu/error-report.h" #include "qemu/timer.h" #include "qemu/bitops.h" #include "qemu/error-report.h" @@ -399,13 +400,18 @@ static void pflash_update(PFlashCFI01 *pfl, int offset, int size) {int offset_end; +int ret;if (pfl->blk) { offset_end = offset + size;/* widen to sector boundaries */ offset =QEMU_ALIGN_DOWN (offset,BDRV_SECTOR_SIZE ); offset_end =QEMU_ALIGN_UP (offset_end,BDRV_SECTOR_SIZE ); -blk_pwrite (pfl->blk, offset, pfl->storage + offset, + ret =blk_pwrite (pfl->blk, offset, pfl->storage + offset, offset_end - offset,0 ); +if (ret <0 ) { +/* TODO set error bit in status */ +error_report ("Could not update PFLASH: %s" ,strerror (-ret)); + } } }hw/block/pflash_cfi02.c +7
-2@@ -37,6 +37,7 @@#include "hw/block/flash.h" #include "hw/qdev-properties.h" #include "qapi/error.h" +#include "qemu/error-report.h" #include "qemu/bitmap.h" #include "qemu/timer.h" #include "sysemu/block-backend.h" @@ -393,13 +394,18 @@ static uint64_t pflash_read(void *opaque, hwaddr offset, unsigned int width)static void pflash_update (PFlashCFI02 *pfl,int offset,int size) {int offset_end; +int ret;if (pfl->blk) { offset_end = offset + size;/* widen to sector boundaries */ offset =QEMU_ALIGN_DOWN (offset,BDRV_SECTOR_SIZE ); offset_end =QEMU_ALIGN_UP (offset_end,BDRV_SECTOR_SIZE ); -blk_pwrite (pfl->blk, offset, pfl->storage + offset, + ret =blk_pwrite (pfl->blk, offset, pfl->storage + offset, offset_end - offset,0 ); +if (ret <0 ) { +/* TODO set error bit in status */ +error_report ("Could not update PFLASH: %s" ,strerror (-ret)); + } } } --2.21.1
Re: [PATCH-for-5.0 10/12] hw/block/pflash: Check return value of blk_pwrite()
In reply to Philippe Mathieu-DaudéQuoted 5 lines · Philippe Mathieu-Daudé
From: Mansour Ahmadi <[email protected]>
When updating the PFLASH file contents, we should check for a possible failure of
blk_pwrite(). Similar to commit3a688294e.
There is actually a Coverity report for this issue, CID 1357678 (Unchecked return value) from 2016-07-15 06:28:48:
CID 1357678 (#2 of 2): Unchecked return value (CHECKED_RETURN). check_return: Calling blk_pwrite without checking return value (as is done elsewhere 52 out of 59 times).
So it seems fair to add:
Quoted 74 lines
Signed-off-by Mansour Ahmadi <[email protected]>Message-Id: <[email protected]>
[PMD: Add missing "qemu/error-report.h" include and TODO comment]Signed-off-by Philippe Mathieu-Daudé <[email protected]>--- hw/block/pflash_cfi01.c | 8 +++++++- hw/block/pflash_cfi02.c | 8 +++++++- 2 files changed, 14 insertions(+), 2 deletions(-)hw/block/pflash_cfi01.c +7
-1@@ -42,6 +42,7 @@#include "hw/qdev-properties.h" #include "sysemu/block-backend.h" #include "qapi/error.h" +#include "qemu/error-report.h" #include "qemu/timer.h" #include "qemu/bitops.h" #include "qemu/error-report.h" @@ -399,13 +400,18 @@ static void pflash_update(PFlashCFI01 *pfl, int offset, int size) {int offset_end; +int ret;if (pfl->blk) { offset_end = offset + size;/* widen to sector boundaries */ offset =QEMU_ALIGN_DOWN (offset,BDRV_SECTOR_SIZE ); offset_end =QEMU_ALIGN_UP (offset_end,BDRV_SECTOR_SIZE ); -blk_pwrite (pfl->blk, offset, pfl->storage + offset, + ret =blk_pwrite (pfl->blk, offset, pfl->storage + offset, offset_end - offset,0 ); +if (ret <0 ) { +/* TODO set error bit in status */ +error_report ("Could not update PFLASH: %s" ,strerror (-ret)); + } } }hw/block/pflash_cfi02.c +7
-1@@ -37,6 +37,7 @@#include "hw/block/flash.h" #include "hw/qdev-properties.h" #include "qapi/error.h" +#include "qemu/error-report.h" #include "qemu/bitmap.h" #include "qemu/timer.h" #include "sysemu/block-backend.h" @@ -393,13 +394,18 @@ static uint64_t pflash_read(void *opaque, hwaddr offset, unsigned int width)static void pflash_update (PFlashCFI02 *pfl,int offset,int size) {int offset_end; +int ret;if (pfl->blk) { offset_end = offset + size;/* widen to sector boundaries */ offset =QEMU_ALIGN_DOWN (offset,BDRV_SECTOR_SIZE ); offset_end =QEMU_ALIGN_UP (offset_end,BDRV_SECTOR_SIZE ); -blk_pwrite (pfl->blk, offset, pfl->storage + offset, + ret =blk_pwrite (pfl->blk, offset, pfl->storage + offset, offset_end - offset,0 ); +if (ret <0 ) { +/* TODO set error bit in status */ +error_report ("Could not update PFLASH: %s" ,strerror (-ret)); + } } }
PATCH 11/12 gdbstub: Do not use memset() on GByteArray
In reply to Philippe Mathieu-DaudéIntroduce gdb_get_zeroes() to fill a GByteArray with zeroes.
---
Since v1: Use memset (pm215)
--- include/exec/gdbstub.h | 10 ++++++++++ target/arm/gdbstub.c | 3 +-- target/xtensa/gdbstub.c | 6 ++---- 3 files changed, 13 insertions(+), 6 deletions(-)
include/exec/gdbstub.h +10 -0
@@ -125,6 +125,16 @@ static inline int gdb_get_reg128(GByteArray *buf, uint64_t val_hi,return 16 ; } +static inline int gdb_get_zeroes (GByteArray *array,size_t len) +{ +guint oldlen = array->len; + +g_byte_array_set_size (array, oldlen + len); +memset (array->data + oldlen,0 , len); + +return len; +} +/** * gdb_get_reg_ptr: get pointer to start of last element * @len: length of element
target/arm/gdbstub.c +1 -2
@@ -47,8 +47,7 @@ int arm_cpu_gdb_read_register(CPUState *cs, GByteArray *mem_buf, int n)if (gdb_has_xml) {return 0 ; } -memset (mem_buf,0 ,12 ); -return 12 ; +return gdb_get_zeroes (mem_buf,12 ); }switch (n) {case 24 :
target/xtensa/gdbstub.c +2 -4
@@ -105,8 +105,7 @@ int xtensa_cpu_gdb_read_register(CPUState *cs, GByteArray *mem_buf, int n)default :qemu_log_mask (LOG_UNIMP ,"%s from reg %d of unsupported size %d\n" , __func__, n, reg->size); -memset (mem_buf,0 , reg->size); -return reg->size; +return gdb_get_zeroes (mem_buf, reg->size); }case xtRegisterTypeWindow:/*a*/
@@ -115,8 +114,7 @@ int xtensa_cpu_gdb_read_register(CPUState *cs, GByteArray *mem_buf, int n)default :qemu_log_mask (LOG_UNIMP ,"%s from reg %d of unsupported type %d\n" , __func__, n, reg->type); -memset (mem_buf,0 , reg->size); -return reg->size; +return gdb_get_zeroes (mem_buf, reg->size); } }
Signature · 1 line
2.21.1
PATCH 12/12 gdbstub: Introduce gdb_get_freg32() to get float32 registers
In reply to Philippe Mathieu-DaudéSince we now use a GByteArray, we can not use stfl_p() directly. Introduce the gdb_get_freg32() helper to load a float32 register.
--- include/exec/gdbstub.h | 12 ++++++++++++ target/sh4/gdbstub.c | 6 ++---- 2 files changed, 14 insertions(+), 4 deletions(-)
include/exec/gdbstub.h +12 -0
@@ -125,6 +125,18 @@ static inline int gdb_get_reg128(GByteArray *buf, uint64_t val_hi,return 16 ; } +static inline int gdb_get_freg32 (GByteArray *array,float32 val) +{ +uint8_t buf[4 ]; + +QEMU_BUILD_BUG_ON (sizeof (CPU_FloatU) !=sizeof (buf)); + +stfl_p (buf, val); +g_byte_array_append (array, buf,sizeof (buf)); + +return sizeof (buf); +} +static inline int gdb_get_zeroes (GByteArray *array,size_t len) {guint oldlen = array->len;
target/sh4/gdbstub.c +2 -4
@@ -58,11 +58,9 @@ int superh_cpu_gdb_read_register(CPUState *cs, GByteArray *mem_buf, int n)return gdb_get_regl (mem_buf, env->fpscr); case 25 ... 40:if (env->fpscr &FPSCR_FR ) { -stfl_p (mem_buf, env->fregs[n -9 ]); - }else { -stfl_p (mem_buf, env->fregs[n -25 ]); +return gdb_get_freg32 (mem_buf, env->fregs[n -9 ]); } -return 4 ; +return gdb_get_freg32 (mem_buf, env->fregs[n -25 ]);case 41 :return gdb_get_regl (mem_buf, env->ssr);case 42 :
Signature · 1 line
2.21.1
Quoted 6 lines · Philippe Mathieu-Daudé
Collection of bugfixes for 5.0.
Only vhost-user-gpu/grlib_ahb_apb_pnp patches are reviewed.
As 5.0-rc3 is tomorrow, I thought it could help to gather them and resend altogether.
So who's applying all this stuff? Peter?
Quoted 42 lines
Regards,
Phil.
Mansour Ahmadi (1):
hw/block/pflash: Check return value ofblk_pwrite()Philippe Mathieu-Daudé (11):
Revert "prevent crash when executing guest-file-read with large count"
qga: Extractguest_file_handle_find()tocommands-common.h
qga: Extractqmp_guest_file_read()to commoncommands.c
qga: Restrict guest-file-read count to 48 MB to avoid crashes
vhost-user-gpu: Release memory returned byvu_queue_pop()withfree()hw/openrisc/pic_cpu: Use qdev gpio rather thanqemu_allocate_irqs()hw/misc/grlib_ahb_apb_pnp: Avoid crash when writing to AHB PnP
registers
hw/misc/grlib_ahb_apb_pnp: Fix AHB PnP 8-bit accesses hw/display/sm501: Avoid heap overflow insm501_2d_operation()
gdbstub: Do not usememset()on GByteArray
gdbstub: Introducegdb_get_freg32()to get float32 registersqga/qapi-schema.json | 6 +++-- include/exec/gdbstub.h | 22 ++++++++++++++++ qga/commands-common.h | 21 +++++++++++++++ contrib/vhost-user-gpu/vhost-user-gpu.c | 4 +-- contrib/vhost-user-gpu/virgl.c | 2 +- hw/block/pflash_cfi01.c | 8 +++++- hw/block/pflash_cfi02.c | 8 +++++- hw/display/sm501.c | 6 +++++ hw/misc/grlib_ahb_apb_pnp.c | 11 ++++++++ hw/openrisc/pic_cpu.c | 5 ++-- qga/commands-posix.c | 29 +++++--------------- qga/commands-win32.c | 35 ++++++------------------- qga/commands.c | 33 +++++++++++++++++++++++ target/arm/gdbstub.c | 3 +-- target/sh4/gdbstub.c | 6 ++--- target/xtensa/gdbstub.c | 6 ++--- 16 files changed, 136 insertions(+), 69 deletions(-) create mode 100644 qga/commands-common.hSignature · 1 line
2.21.1
Quoted 11 lines · Michael S. Tsirkin
Quoted 6 lines · Philippe Mathieu-Daudé
Collection of bugfixes for 5.0.
Only vhost-user-gpu/grlib_ahb_apb_pnp patches are reviewed.
As 5.0-rc3 is tomorrow, I thought it could help to gather them and resend altogether.
So who's applying all this stuff? Peter?
I talked to Philippe about this on irc; I cherry picked the one bug that we thought was critical for 5.0 (the vhost-user-gpu free/g_free mismatch bugfix) and the rest will be for 5.1.
thanks
-- PMM