hw/scsi/esp.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-)
Following on from the ESP changes in QEMU 6.0 someone pointed out that the old Linux 2.6 ESP driver as used in Aurelian's SPARC image at https://people.debian.org/~aurel32/qemu/sparc/ emits a constant stream of "esp0: STEP_ASEL for tgt 0" messages to the console during boot. These patches solve the issue so that the older image is able to boot cleanly once again. The first patch is a genuine bug fix for the QEMU 6.0 changes whilst the second works around the delayed bus phase change when deferring IO transfers from the end of the command phase to the start of the information transfer phase. I've tested this using the extended suite of images used for the QEMU 6.0 changes and confirmed that there are no other regressions. Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> Mark Cave-Ayland (2): esp: only assert INTR_DC interrupt flag if selection fails esp: only set ESP_RSEQ at the start of the select sequence hw/scsi/esp.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) -- 2.20.1
On 18/05/21 23:25, Mark Cave-Ayland wrote: > Following on from the ESP changes in QEMU 6.0 someone pointed out that the old > Linux 2.6 ESP driver as used in Aurelian's SPARC image at > https://people.debian.org/~aurel32/qemu/sparc/ emits a constant stream of > "esp0: STEP_ASEL for tgt 0" messages to the console during boot. > > These patches solve the issue so that the older image is able to boot cleanly > once again. The first patch is a genuine bug fix for the QEMU 6.0 changes whilst > the second works around the delayed bus phase change when deferring IO transfers > from the end of the command phase to the start of the information transfer > phase. > > I've tested this using the extended suite of images used for the QEMU 6.0 changes > and confirmed that there are no other regressions. > > Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> > > > Mark Cave-Ayland (2): > esp: only assert INTR_DC interrupt flag if selection fails > esp: only set ESP_RSEQ at the start of the select sequence > > hw/scsi/esp.c | 13 ++++++++++--- > 1 file changed, 10 insertions(+), 3 deletions(-) > Queued, thanks. Paolo
On 18/05/2021 22:25, Mark Cave-Ayland wrote: > Following on from the ESP changes in QEMU 6.0 someone pointed out that the old > Linux 2.6 ESP driver as used in Aurelian's SPARC image at > https://people.debian.org/~aurel32/qemu/sparc/ emits a constant stream of > "esp0: STEP_ASEL for tgt 0" messages to the console during boot. > > These patches solve the issue so that the older image is able to boot cleanly > once again. The first patch is a genuine bug fix for the QEMU 6.0 changes whilst > the second works around the delayed bus phase change when deferring IO transfers > from the end of the command phase to the start of the information transfer > phase. > > I've tested this using the extended suite of images used for the QEMU 6.0 changes > and confirmed that there are no other regressions. > > Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> > > > Mark Cave-Ayland (2): > esp: only assert INTR_DC interrupt flag if selection fails > esp: only set ESP_RSEQ at the start of the select sequence > > hw/scsi/esp.c | 13 ++++++++++--- > 1 file changed, 10 insertions(+), 3 deletions(-) Ping? ATB, Mark.
On 28/05/2021 08:07, Mark Cave-Ayland wrote: > On 18/05/2021 22:25, Mark Cave-Ayland wrote: > >> Following on from the ESP changes in QEMU 6.0 someone pointed out that the old >> Linux 2.6 ESP driver as used in Aurelian's SPARC image at >> https://people.debian.org/~aurel32/qemu/sparc/ emits a constant stream of >> "esp0: STEP_ASEL for tgt 0" messages to the console during boot. >> >> These patches solve the issue so that the older image is able to boot cleanly >> once again. The first patch is a genuine bug fix for the QEMU 6.0 changes whilst >> the second works around the delayed bus phase change when deferring IO transfers >> from the end of the command phase to the start of the information transfer >> phase. >> >> I've tested this using the extended suite of images used for the QEMU 6.0 changes >> and confirmed that there are no other regressions. >> >> Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> >> >> >> Mark Cave-Ayland (2): >> esp: only assert INTR_DC interrupt flag if selection fails >> esp: only set ESP_RSEQ at the start of the select sequence >> >> hw/scsi/esp.c | 13 ++++++++++--- >> 1 file changed, 10 insertions(+), 3 deletions(-) > > Ping? Ping again? ATB, Mark.
© 2016 - 2024 Red Hat, Inc.