| Message ID | 1475675109-8105-1-git-send-email-quintela@redhat.com |
|---|---|
| State | New |
| Headers | show |
On 5 October 2016 at 14:44, Juan Quintela <quintela@redhat.com> wrote: > From: Juan Quintela <quintela@trasno.org> > > The following changes since commit bbc4c3f4f3c624e2de64fdcb79f4dd8c1a508e9d: > > Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' into staging (2016-10-04 14:25:08 +0100) > > are available in the git repository at: > > git://github.com/juanquintela/qemu.git tags/migration/20161005-1 > > for you to fetch changes up to c62da143b6792cc32dbd5db15b936d7f58cc36ee: > > docs/xbzrle: correction (2016-10-05 14:28:41 +0200) > > ---------------------------------------------------------------- > migration/next for 20161005 > > Hi > > This get: > - documentation fix: Cao > - improve max-bandwidth and downtime-limit (Ashijeet) > - move migration commands to "boxed" (Eric) > - rdma fixes (David) > - postcopy fixes (David) > - better errors (David) > > Please apply This runs into the "OSX %zu vs SIZE_MAX" issue again: /Users/pm215/src/qemu-for-merges/migration/migration.c:810:61: warning: format specifies type 'size_t' (aka 'unsigned long') but the argument has type 'unsigned long long' [-Wformat] " range of 0 to %zu bytes/second", SIZE_MAX); ~~~ ^~~~~~~~ %llu /Users/pm215/src/qemu-for-merges/include/qapi/error.h:163:35: note: expanded from macro 'error_setg' (fmt), ## __VA_ARGS__) ^ /usr/include/stdint.h:153:20: note: expanded from macro 'SIZE_MAX' #define SIZE_MAX UINT64_MAX ^~~~~~~~~~ /usr/include/stdint.h:87:27: note: expanded from macro 'UINT64_MAX' #define UINT64_MAX 18446744073709551615ULL ^~~~~~~~~~~~~~~~~~~~~~~ In the absence of a suitable configure test and override of SIZE_MAX, we've been adding explicit (uint64_t) casts in the printf as the simplest workaround. thanks -- PMM
Peter Maydell <peter.maydell@linaro.org> wrote: > On 5 October 2016 at 14:44, Juan Quintela <quintela@redhat.com> wrote: >> From: Juan Quintela <quintela@trasno.org> >> >> The following changes since commit bbc4c3f4f3c624e2de64fdcb79f4dd8c1a508e9d: >> >> Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' >> into staging (2016-10-04 14:25:08 +0100) >> >> are available in the git repository at: >> >> git://github.com/juanquintela/qemu.git tags/migration/20161005-1 >> >> for you to fetch changes up to c62da143b6792cc32dbd5db15b936d7f58cc36ee: >> >> docs/xbzrle: correction (2016-10-05 14:28:41 +0200) >> >> ---------------------------------------------------------------- >> migration/next for 20161005 >> >> Hi >> >> This get: >> - documentation fix: Cao >> - improve max-bandwidth and downtime-limit (Ashijeet) >> - move migration commands to "boxed" (Eric) >> - rdma fixes (David) >> - postcopy fixes (David) >> - better errors (David) >> >> Please apply > > This runs into the "OSX %zu vs SIZE_MAX" issue again: Grrrr. I already have trouble with %PR vs %zu for i386 vs anything else. Anyone has a linux crosscompiler for osx that I can use? Or there is a way to get darwin or whatever virtual machine to test compile this? It gets frustrating, before submissing I do: - make check - avocado virt test - full compile on linux x86_64 (64 bits) - full compile on linux i386 (32 bits) - full cross-compile for windows 64bit - full cross-compile for windows 32bit (ok, I only test linux 64 bit), and I still don't get all the errors. > /Users/pm215/src/qemu-for-merges/migration/migration.c:810:61: > warning: format specifies type 'size_t' (aka 'unsigned long') but the > argument has type 'unsigned long long' [-Wformat] > " range of 0 to %zu bytes/second", SIZE_MAX); > ~~~ ^~~~~~~~ > %llu > /Users/pm215/src/qemu-for-merges/include/qapi/error.h:163:35: note: > expanded from macro 'error_setg' > (fmt), ## __VA_ARGS__) > ^ > /usr/include/stdint.h:153:20: note: expanded from macro 'SIZE_MAX' > #define SIZE_MAX UINT64_MAX > ^~~~~~~~~~ > /usr/include/stdint.h:87:27: note: expanded from macro 'UINT64_MAX' > #define UINT64_MAX 18446744073709551615ULL > ^~~~~~~~~~~~~~~~~~~~~~~ > > In the absence of a suitable configure test and override of SIZE_MAX, > we've been adding explicit (uint64_t) casts in the printf as the > simplest workaround. Will do that, but as I have no way to compile test this, it will continue hitting me :-( Thanks for telling me. Later, Juan. PD. Sorry if it sounded like a rant, just frustration O:-)
On 6 October 2016 at 17:09, Juan Quintela <quintela@redhat.com> wrote: > Peter Maydell <peter.maydell@linaro.org> wrote: >> On 5 October 2016 at 14:44, Juan Quintela <quintela@redhat.com> wrote: >>> From: Juan Quintela <quintela@trasno.org> >>> >>> The following changes since commit bbc4c3f4f3c624e2de64fdcb79f4dd8c1a508e9d: >>> >>> Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' >>> into staging (2016-10-04 14:25:08 +0100) >>> >>> are available in the git repository at: >>> >>> git://github.com/juanquintela/qemu.git tags/migration/20161005-1 >>> >>> for you to fetch changes up to c62da143b6792cc32dbd5db15b936d7f58cc36ee: >>> >>> docs/xbzrle: correction (2016-10-05 14:28:41 +0200) >>> >>> ---------------------------------------------------------------- >>> migration/next for 20161005 >>> >>> Hi >>> >>> This get: >>> - documentation fix: Cao >>> - improve max-bandwidth and downtime-limit (Ashijeet) >>> - move migration commands to "boxed" (Eric) >>> - rdma fixes (David) >>> - postcopy fixes (David) >>> - better errors (David) >>> >>> Please apply >> >> This runs into the "OSX %zu vs SIZE_MAX" issue again: > > Grrrr. > I already have trouble with %PR vs %zu for i386 vs anything else. > > Anyone has a linux crosscompiler for osx that I can use? Or there is a > way to get darwin or whatever virtual machine to test compile this? Travis builds include OSX, if you're prepared to wait around for them to complete before submitting (they're usually about 10 hours wall-clock time to complete a build). > It gets frustrating, before submissing I do: > - make check > - avocado virt test > - full compile on linux x86_64 (64 bits) > - full compile on linux i386 (32 bits) > - full cross-compile for windows 64bit > - full cross-compile for windows 32bit > > (ok, I only test linux 64 bit), and I still don't get all the errors. This is why I do OSX builds in my merge tests, they do pick up some stuff other builds don't. thanks -- PMM
Peter Maydell <peter.maydell@linaro.org> wrote: > On 6 October 2016 at 17:09, Juan Quintela <quintela@redhat.com> wrote: >> Peter Maydell <peter.maydell@linaro.org> wrote: >>> On 5 October 2016 at 14:44, Juan Quintela <quintela@redhat.com> wrote: >>>> From: Juan Quintela <quintela@trasno.org> >>>> >>>> The following changes since commit bbc4c3f4f3c624e2de64fdcb79f4dd8c1a508e9d: >>>> >>>> Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' >>>> into staging (2016-10-04 14:25:08 +0100) >>>> >>>> are available in the git repository at: >>>> >>>> git://github.com/juanquintela/qemu.git tags/migration/20161005-1 >>>> >>>> for you to fetch changes up to c62da143b6792cc32dbd5db15b936d7f58cc36ee: >>>> >>>> docs/xbzrle: correction (2016-10-05 14:28:41 +0200) >>>> >>>> ---------------------------------------------------------------- >>>> migration/next for 20161005 >>>> >>>> Hi >>>> >>>> This get: >>>> - documentation fix: Cao >>>> - improve max-bandwidth and downtime-limit (Ashijeet) >>>> - move migration commands to "boxed" (Eric) >>>> - rdma fixes (David) >>>> - postcopy fixes (David) >>>> - better errors (David) >>>> >>>> Please apply >>> >>> This runs into the "OSX %zu vs SIZE_MAX" issue again: >> >> Grrrr. >> I already have trouble with %PR vs %zu for i386 vs anything else. >> >> Anyone has a linux crosscompiler for osx that I can use? Or there is a >> way to get darwin or whatever virtual machine to test compile this? > > Travis builds include OSX, if you're prepared to wait around > for them to complete before submitting (they're usually about 10 > hours wall-clock time to complete a build). And I was complaining that each of my full builds after a rebase can take around 45mins ..... O:-) They are all local. > >> It gets frustrating, before submissing I do: >> - make check >> - avocado virt test >> - full compile on linux x86_64 (64 bits) >> - full compile on linux i386 (32 bits) >> - full cross-compile for windows 64bit >> - full cross-compile for windows 32bit >> >> (ok, I only test linux 64 bit), and I still don't get all the errors. > > This is why I do OSX builds in my merge tests, they do pick > up some stuff other builds don't. > > thanks > -- PMM Later, Juan.
On 6 October 2016 at 17:23, Juan Quintela <quintela@redhat.com> wrote: > Peter Maydell <peter.maydell@linaro.org> wrote: >> Travis builds include OSX, if you're prepared to wait around >> for them to complete before submitting (they're usually about 10 >> hours wall-clock time to complete a build). > > And I was complaining that each of my full builds after a rebase can > take around 45mins ..... O:-) They are all local. None of the individual travis jobs take very long, I think it just throttles our access to infrastructure and refuses to parallelize us very much. thanks -- PMM
On 6 October 2016 at 11:59, Peter Maydell <peter.maydell@linaro.org> wrote: > On 5 October 2016 at 14:44, Juan Quintela <quintela@redhat.com> wrote: >> From: Juan Quintela <quintela@trasno.org> >> >> The following changes since commit bbc4c3f4f3c624e2de64fdcb79f4dd8c1a508e9d: >> >> Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' into staging (2016-10-04 14:25:08 +0100) >> >> are available in the git repository at: >> >> git://github.com/juanquintela/qemu.git tags/migration/20161005-1 >> >> for you to fetch changes up to c62da143b6792cc32dbd5db15b936d7f58cc36ee: >> >> docs/xbzrle: correction (2016-10-05 14:28:41 +0200) >> >> ---------------------------------------------------------------- >> migration/next for 20161005 >> >> Hi >> >> This get: >> - documentation fix: Cao >> - improve max-bandwidth and downtime-limit (Ashijeet) >> - move migration commands to "boxed" (Eric) >> - rdma fixes (David) >> - postcopy fixes (David) >> - better errors (David) >> >> Please apply > > This runs into the "OSX %zu vs SIZE_MAX" issue again Since Eric has now kindly provided a proper configure test and generic workaround for the SIZE_MAX bug and it is in master, I tried just re-applying this pull request, but unfortunately it has a merge conflict in qapi-schema.json that I don't understand well enough to resolve. So if you could produce a new pull request that would be best, but you don't need to worry about the OSX failure any more. thanks -- PMM
From: Juan Quintela <quintela@trasno.org> The following changes since commit bbc4c3f4f3c624e2de64fdcb79f4dd8c1a508e9d: Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' into staging (2016-10-04 14:25:08 +0100) are available in the git repository at: git://github.com/juanquintela/qemu.git tags/migration/20161005-1 for you to fetch changes up to c62da143b6792cc32dbd5db15b936d7f58cc36ee: docs/xbzrle: correction (2016-10-05 14:28:41 +0200) ---------------------------------------------------------------- migration/next for 20161005 Hi This get: - documentation fix: Cao - improve max-bandwidth and downtime-limit (Ashijeet) - move migration commands to "boxed" (Eric) - rdma fixes (David) - postcopy fixes (David) - better errors (David) Please apply Thanks, Juan ---------------------------------------------------------------- Ashijeet Acharya (2): migrate: Fix bounds check for migration parameters in migration.c migrate: move max-bandwidth and downtime-limit to migrate_set_parameter Cao jin (1): docs/xbzrle: correction Dr. David Alan Gilbert (9): migration: report an error giving the failed field migration: Report values for comparisons migration/rdma: Pass qemu_file errors across link migration: Make failed migration load set file error migration/rdma: Don't flag an error when we've been told about one Postcopy vs xbzrle: Don't send xbzrle pages once in postcopy [for 2.8] RAMBlocks: Store page size migration/postcopy: Explicitly disallow huge pages migration: Fix seg with missing port Eric Blake (3): migrate: Fix cpu-throttle-increment regression in HMP migrate: Share common MigrationParameters struct migrate: Use boxed qapi for migrate-set-parameters docs/qmp-commands.txt | 13 +++- docs/xbzrle.txt | 4 +- exec.c | 19 +++-- hmp.c | 77 ++++++++++++++------- include/exec/cpu-common.h | 1 + include/exec/ram_addr.h | 1 + include/migration/migration.h | 1 - migration/migration.c | 157 +++++++++++++++++++++++------------------- migration/postcopy-ram.c | 24 +++++++ migration/ram.c | 4 +- migration/rdma.c | 9 ++- migration/savevm.c | 19 +++-- migration/socket.c | 16 +++-- migration/vmstate.c | 10 +++ qapi-schema.json | 109 +++++++++++++---------------- 15 files changed, 282 insertions(+), 182 deletions(-)